Translational Advances in the Management of Acute Spinal Cord Injury

Abstract

AANS: American Association of Neurological Surgeons AIS: Abbreviated Injury Scale ASIA: American Spinal Injury Association CNS: Congress of Neurological Surgeons CSFD: cerebrospinal fluid drainage FDA: Food and Drug Administration FGF: fibroblast growth factor G-CSF: granulocyte colony-stimulating factor HGF: hepatocyte growth factor IL: interleukin iPSC: induced pluripotent stem cell MAP: mean arterial blood pressure MPSS: methylprednisolone sodium succinate MSC: mesenchymal stem cell NASCIS: National Acute Spinal Cord Injury Studies Nogo: neurite outgrowth inhibitor NgR: Nogo receptor NPC: neural precursor cell NSS: Neuro-Spinal Scaffold OEC: olfactory ensheathing cell OPC: oligodendrocyte progenitor cell PEG: polyethylene glycol PLGA: poly(lactic-co-glycolic acid) SCI: spinal cord injury TH: therapeutic hypothermia TNF: tumor necrosis factor Traumatic spinal cord injury (SCI) is a devastating event caused by trauma to the spine which leads to mechanical disruption of the spinal cord. The incidence of SCI varies worldwide. Focusing on developed regions, North America (39 per million) has a higher annual incidence than Australia (16 per million) or Western Europe (15 per million).1 Direct costs for lifetime patient care reach $1.1 to 4.6 million per patient, which further underscores the need for the development of effective SCI treatments.2 Substantial research effort has been dedicated to uncovering the pathophysiology of SCI. This has led to the development of pharmacologic and cell-based therapies, which are now demonstrating functional motor recovery in animal models. Among these, several promising therapeutic agents are already being investigated in clinical trials for SCI. This review will summarize the pathophysiology and current evidence-based clinical strategies to manage an acute spinal cord injury followed by a discussion of key emerging treatments including pharmacological approaches, cell-based therapies, biomaterials and physiological approaches. PATHOPHYSIOLOGY Phases of SCI Tissue damage after SCI has been divided into primary and secondary injury phases.3,4 The physical forces of the initial trauma cause the primary injury and this is the main determinant of the severity of SCI. The axons, blood vessels, and cell membranes are disrupted by physical forces such as compression, shearing, laceration, and acute stretch. Secondary injury refers to delayed, progressive damage which continues after the primary injury and represents an additional important determinant of neurological deficits (Figure).5,6 Due to the disruption of the blood–spinal cord barrier following the primary injury, infiltration of inflammatory cells such as macrophages, microglia, T-cells, and neutrophils can be observed. Inflammatory cytokines such as tumor necrosis factor (TNF) α, interleukin (IL)-1α, IL-1β, and IL-6 are released by these cells, with levels of these cytokines peaking 6 to 12 h after injury and remaining elevated up to 4 d after injury.7 Increases in intracellular calcium are caused by the disruption of ionic homeostasis after SCI and activates calcium-dependent proteases (eg, phospholipases, calpain, caspase, and nitric oxide synthase). These proteases trigger dysfunction of mitochondria which leads to cell death.8 Oligodendrocytes are highly susceptible to apoptotic loss and apoptosis has been observed, not only at the lesion epicenter, but also distant from the epicenter leading to demyelination of preserved axons.9-11 Furthermore, delayed necrosis and apoptosis are induced by reactive oxygen species which are released by phagocytic inflammatory cells.12-14 Moreover, the disrupted cells release excitatory amino acids (eg, glutamate and asparate) after SCI15,16 and the excessive activation of excitatory amino acid receptors causes further loss of neurons and glia by both necrotic and apoptotic cell death.17 To achieve repair and regeneration of the injured spinal cord, researchers have attempted to disrupt elements of the secondary injury pathway with the aim of neural preservation, inhibition of the barriers to axonal regeneration, and replacement of the damaged cells by cell transplantation therapy. From a pathophysiological perspective, it is likely that the optimal therapy will be a combinatorial one consisting of administration of drugs to reduce secondary injury at the acute phase, followed by cell transplantation or other regenerative therapies to regenerate the damaged spinal cord tissue in the subacute to chronic phases.18,19 These therapies are discussed in greater detail below.FIGURE: Three pathophysiological phases after SCI including acute (eg, hemorrhage, edema, and inflammation), subacute (eg, demyelination and axonal dieback), and chronic (eg, cavity formation) phases. Primary injury is caused by the physical forces of the initial traumatic event. Secondary injury refers to delayed, progressive damage which includes inflammation, loss of ionic homeostasis, oxidative damage, excitotoxicity, apoptosis, and necrosis. Oligodendrocytes are highly susceptible to apoptotic loss resulting in axonal demyelination. Cystic cavitation forms in the center of the spinal cord, with surrounding glial scar in the subacute and chronic phases. Nonastrocyte cells mainly form a chemical barrier by secreting growth inhibitory CSPGs.Barriers to Regeneration The adult mammalian Congress of Neurological Surgeons (CNS), including the spinal cord, has generally been considered to have limited regenerative capacity due to the finite number of available regenerative cells and the restricted plasticity of the adult CNS.20 While recent research has shown that the spinal cord has more regenerative capacity than was previously thought,21,22 compared with the peripheral nervous system, the regenerative capacity of the CNS is lower and it gradually decreases with increasing age.23 Schwab et al24 reported the inhibitory nature of CNS myelin in 1985. Myelin-associated proteins, such as neurite outgrowth inhibitor A (Nogo A),25,26 oligodendrocyte-myelin glycoprotein,27 and myelin-associated glycoprotein28,29 function through Nogo receptors (NgR). The NgRs lack an intracellular signaling domain and transduce inhibitory signals by forming coreceptor complexes with TNF receptor family proteins (eg, p75, TROY, and LIGO-1) to activate the GTPase Rho A. The downstream effector of Rho A is Rho-associated protein kinase which affects changes in the actin cytoskeleton and leads to growth cone collapse of regenerating axons, neurite retraction, and increasing apoptosis. SCI is accompanied by mechanically induced and excitotoxic cell death, with associated demyelination. The lost parenchyma is replaced by cystic cavitation and regeneration is often hindered by the presence of this cystic cavity which lacks the substrate to support axonal growth and cell migration.30 Furthermore, at the site of injury, glial and fibrotic scarring is also present (Figure). Glial and fibrotic scarring results when pericytes, hypertrophied astrocytes, fibroblast lineage cells, and inflammatory cells form a physical barrier, walling off injured tissue from healthy tissue.31,32 Recent research has shown that both astrocytes and nonastrocyte cells can form a physical and chemical barrier by secreting growth inhibitory chondroitin sulfate proteoglycans (CSPGs) such as neurocan, versican, brevican, phosphacan, and NG2.33 A fibroblast-derived scar can also be located in the perilesional region and is associated with the deposition of inhibitory extracellular matrix molecules. Similar to myelin-associated inhibitors, these molecules act as chemical barriers to the regeneration of axons. CURRENT CLINICAL STRATEGIES Early Surgical Intervention To reduce the effects of cord compression and resultant ischemia, early bony and ligamentous surgical decompression is performed to provide relief from the mechanical pressure. To elucidate the effectiveness of early decompression, a prospective cohort study, The Surgical Treatment of Acute Spinal Cord Injury Study (STASCIS) was conducted with 313 cervical SCI patients.34 After adjusting for confounders, the early decompression group (<24 h after SCI) was 2.8 times as likely to demonstrate an Abbreviated Injury Scale (AIS) improvement of 2 or more grades at 6 mo after SCI compared with the late decompression group (≥24 h after SCI). A subsequent prospective Canadian cohort study (including cervical, thoracic, and lumbar SCI, n = 84) also revealed that early decompression was associated with a 2 or more grade AIS improvement at the time of rehabilitation facility discharge.35 The findings of these studies support the concept of “Time is Spine” which emphasizes the importance of early diagnosis and intervention to improve long-term outcomes. Central Cord Syndrome Central cord injury is characterized by greater weakness in the upper extremities than the lower extremities, variable sensory loss, variable bowel/bladder dysfunction, and, usually, early rapid improvements in neurological function. Early decompression has traditionally been avoided in cases of central cord injury with patients being allowed to plateau in their recovery over a number of weeks before any intervention.36 However, for patients with pre-existing canal stenosis, recent evidence suggests that early surgery may improve long-term outcomes. A systematic review demonstrated that patients undergoing early decompression (<24 h after SCI) had American Spinal Injury Association (ASIA) motor scores that were 6.31 points higher, and a greater chance of improvement in ASIA grade (odds ratio of 2.81) at 12-mo follow-up than those undergoing late decompression (≥24 h after SCI).37 Although the prospective randomized controlled Comparing Surgical Decompression Versus Conservative Treatment in Incomplete Spinal Cord Injury (COSMIC, NCT01367405) trial was initiated in 2013, it was terminated in 2016 due to difficulties in enrolling patients. Blood Pressure Augmentation The neuroprotective effects of blood pressure augmentation act through enhancing systemic perfusion. Several studies have shown that high-normal mean arterial blood pressures (MAPs) of 85 to 90 mm Hg may improve outcomes in SCI patients.38-40 The guidelines of the American Association of Neurological Surgeons (AANS) and (CNS) recommend MAP targets of 85 to 90 mm Hg as an option in SCI to be initiated as early as possible and maintained for 7 d after injury.41 This MAP elevation requires invasive blood pressure monitoring, maintenance of slightly hypervolemic state, and central venous access for continuous infusion of vasopressors. A noninferiority trial named Mean Arterial Blood Pressure Treatment for Acute Spinal Cord Injury (MAPS; NCT02232165) comparing MAP ≥ 85 mm Hg and MAP ≥ 65 mm Hg has been developed to assess the efficacy of lower targets. ASIA motor scores at 1 yr postinjury will be evaluated, and this trial is expected to complete in March 2017. Steroids for SCI Methylprednisolone sodium succinate (MPSS) is the only agent from completed clinical trials that has entered clinical use. It acts by reducing oxidative stress to enhance neural cell survival in animal models of traumatic SCI. Three landmark National Acute Spinal Cord Injury Studies (NASCIS) examined the use of MPSS for acute SCI.42-47 Although no neurological benefit in the MPSS-treated group was observed in the overall analyses of these studies, a subgroup analysis in the NASCIS II and III trials demonstrated that use of the drug in a higher dosing regimen than that used in NASCIS I within 8 h of injury resulted in neurological improvement, and that MPSS bolus 3 to 8 h after injury improved neurological function when it was administered for 48 h rather than 24 h.44-47 Recent evidence further supports the use of MPSS for SCI. A 2012 Cochrane meta-analysis and review demonstrated a 4 point greater ASIA motor score improvement in the group that received MPSS for acute SCI and that its administration was not associated with a significant increase in the risk of complications.48 Nevertheless, the 2013 AANS/CNS Section on Disorders of the Spine and Peripheral Nerves guideline provided a level I recommendation against the administration of MPSS which represents a marked change from the previous version despite little change in the evidence considered. Accordingly, an updated AOSpine guideline suggests that 24 h of MPSS IV be administered within 8 h of SCI to patients without medical contraindication.49 Emerging Therapies for SCI Key emerging technologies for SCI treatment include pharmacological approaches, cell-based therapies, biomaterials, and physiological approaches. A summary of these technologies is provided in Table.TABLE: Key Emerging Technologies for Acute SCIPharmacological Approaches Riluzole Riluzole is a benzothiazole antiepileptic which acts via sodium channel blockade. It is approved by the US Food and Drug Administration (FDA), European Medicines Agency, and Health Canada for the treatment of amyotrophic lateral sclerosis.50,51 Its role in neuroprotection stems from its ability to mitigate excitotoxicity and block sodium influx to neurons in addition to restricting the presynaptic release of glutamate.52 In animal studies, Riluzole has been shown to reduce neuronal loss and cavity size which led to improvements in motor function and electrophysiology.53-55 In the phase I trial for acute SCI was recently completed, and 36 patients were enrolled.56 Although elevations of liver enzyme levels were observed temporarily, no serious adverse events were attributed to the drug. Regarding the neurological outcomes, cervical SCI patients treated with riluzole showed the better improvement in ASIA motor score compared with non-riluzole treated patients matched from an historical registry cohort. The phase II/III RCT entitled riluzole in Spinal Cord Injury Study (RISCIS; NCT01597518) is recruiting patients with acute C4-8 injuries with ASIA grade A, B, or C and will compare riluzole versus placebo and assess AIS, Spinal Cord Independence Measure, and brief pain inventory. This study which was initiated in 2014 has to date recruited 70 patients and is expected to conclude in 2020. Minocycline Minocycline is a second-generation semisynthetic tetracycline antibiotic that has the ability to cross the blood–brain barrier. It also has potent anti-inflammatory properties and inhibits microglial activation, TNF-α, IL-1β, cyclooxygenase-2, and matrix metalloproteinases.57-60 In animal studies, minocycline treatment after acute SCI has been shown to protect against neuron loss and reduce the lesion size.61,62 A phase II study showed that patients with incomplete cervical SCI (n = 25) demonstrated an ASIA score improvement of 14 points with minocycline treatment compared to placebo (P = .05).63 The follow-up Phase III Minocycline in Acute Spinal Cord Injury (MASC; NCT01828203) study will compare IV minocycline for 7 d and is expected to conclude in 2018. VX-210 (Cethrin) The Rho pathway is known to negatively impact axonal and neurite growth.64 A toxin produced by Clostridium botulinum, C3 transferase (cethrin), has been shown to inhibit Rho-mediated inhibition of axonal growth which promoted neural regeneration and motor function recovery in rodent SCI models.65 Cethrin is a permeable material intended for application to the dura mater at the site of SCI during decompressive surgery in the acute phase. A phase I/IIa multicenter, dose-escalation human trial evaluating this drug in a human population was published in 201166; no serious adverse events were attributed to the drug.66 Cervical patients treated with 3 mg of cethrin showed improvement in ASIA motor score at 12 mo and this was shown to be superior to historical recovery rates. A phase IIb/III study of cethrin has commenced in cervical SCI patients in 2016 and is expected to conclude in 2018. Anti-Nogo-A antibody (ATI-355) A monoclonal antibody of major inhibitory fractions within CNS myelin, IN-1, has been shown to promote axonal sprouting and functional recovery following SCI in animal models.67 The humanized anti-Nogo antibody, ATI-355, has been shown to promote axonal sprouting and functional recovery following SCI in numerous animal models and is a rare therapeutic in that it has been demonstrated to improve functional outcomes in a primate model.26 A phase I human trial of humanized anti-Nogo antibody (ATI-355) was completed in Europe, rather than the US, as the FDA expressed concerns with the infusion pump. Although this trial has been completed, it has not been published. A phase II study of ATI-355 is about to commence in Europe. Granulocyte Colony Stimulating Factor Granulocyte colony-stimulating factor (G-CSF) has been shown to increase the mobilization of bone marrow stromal cells from the bone marrow and to increase their presence at the site of SCI. In a rodent model, G-CSF enhances neurogenesis, reduces apoptosis, and decreases expression of TNF-α and IL-1β. These positive effects are associated with white matter sparing and improved hind-limb function.68 The phase I/IIa trials, which were nonrandomized, showed no increase in serious adverse events with G-CSF administration alongside AIS grade improvement.69,70 G-CSF is currently in a phases III clinical trial in Japan with results expected in 2018. Hepatocyte Growth Factor Hepatocyte growth factor (HGF) is mainly secreted by mesenchymal cells and promotes cellular growth and motility. HGF enhances neuron survival, decreases lesion size, and reduces oligodendrocyte apoptosis to improve behavioral outcomes in rodent models.71 Moreover, in a primate model of cervical SCI, HGF improved hand dexterity which is one of the most important key functions of the upper limb.72 A phase I/II clinical trial (NCT02193334) comparing intrathecal HGF (KP100IT) versus placebo is now underway with results expected in 2017. Magnesium (AC105) Magnesium is a physiological antagonist of NMDA receptors which decreases excitotoxicity and also functions as an anti-inflammatory agent. Magnesium with polyethylene glycol (PEG) improves cerebrospinal fluid levels without requiring large magnesium doses.73-75 The use of magnesium with PEG in the treatment of animal models of SCI has been shown to enhance tissue sparing and improve motor functional recovery.76,77 However, a phase I/II clinical trial (NCT01750684) of magnesium with PEG (AC105) was terminated in 2015 due to difficulties in enrolling patients. Fibroblast Growth Factor Fibroblast growth factor (FGF) plays a key role in preserving motor neurons adjacent to the SCI site and reduces acute respiratory deficits resulting from the loss of ventral horn neurons by reducing glutamate-mediated excitotoxicity in animal models.78,79 Although a phase I/II trial (NCT01502631) of the FGF-analog (SUN13837) has been completed, the results have not been published to date. Cell-Based Therapies Regenerative therapies based on transplanted multipotent and differentiated cells are an exciting therapeutic approach showing promising results in translational studies. Initial research focused on embryonic stem cell lines derived from aborted early-stage embryos, however, ethical considerations and limited numbers of donor cells created challenges. More recently, the discovery of induced pluripotent stem cells (iPSCs), which can be derived within weeks from any somatic cell source, has revolutionized the field by providing a nearly limitless source of pluripotent cells for research and therapeutic purposes.80 Furthermore, iPSCs can potentially be derived from autologous tissue reducing or eliminating the risk of graft rejection.80 While unforeseen challenges in iPSC technology, such as epigenetic memory and early senescence, have been found, they continue to be a substantial technological advance in spinal cord regeneration.81 The most translationally relevant cell therapies derived from pluripotent stem cells or harvested directed from adult tissue are discussed below. Schwann Cells Schwann cells (SCs) are known to facilitate peripheral nerve regeneration by providing a structural conduit and environmental support to regrowing axons. In rodent models of SCI, grafted SCs have been shown to reduce lesion size, remyelinate axons, and provide increased motor recovery.82 The Miami Project to Cure Paralysis has launched a phase I, open-label trial (n = 10; NCT02354625) to assess SC transplants for patients with chronic AIS grade A-C injuries in the cervical or thoracic spine. The study is expected to conclude in 2018. An additional phase I trial (n = 10; NCT01739023) of autologously derived SCs for AIS grade A thoracic injuries has concluded with results expected in 2017.83 Olfactory Ensheathing Cells Olfactory ensheathing cells (OECs) encircle olfactory neurons and provide protection from bacteria and debris along the CNS-nasal mucosa transition.84-87 In animal models of SCI, they have been found to enhance neurite outgrowth and remyelination resulting in significant functional gains.88 Trials around the world are now exploring autogenic and allogenic OEC transplants for chronic SCI.83 While a meta-analysis of several of these trials (n = 1193) found no increase in serious adverse events, efficacy has yet to be definitively established due to methodological concerns within the studies.89 A previous study showed the bridging effect of transplanted OECs on regenerated axons from the dissected dorsal root into the spinal cord.90 Numerous clinical trials of OECs for chronic SCI have been completed worldwide and analyzed in a meta-analysis which found no significant increase in complication rates related to the transplants.89 Mesenchymal Stem Cells Mesenchymal stem cells (MSCs) are multipotent connective tissue cells capable of differentiating into osteoblasts, chondrocytes, adipocytes, and myocytes to repair musculoskeletal injury.91 Their ability to modulate the local and systemic inflammatory response led to their application in SCI where they were found to promote tissue sparing through neurotrophic paracrine signaling and regulation of inflammation.92-94 Pharmicell Co. (Seoul, South Korea) is now conducting a Phase II/III randomized trial of autologous MSCs delivered via intramedullary and intrathecal injections for patients with AIS grade B cervical SCI within 12 mo of injury. The study is expected to conclude in 2020.83 Neural Precursor Cells Neural precursor cells (NPCs) are multipotent CNS cells capable of differentiating to neurons, astrocytes, and oligodendrocytes to replace lost cells and provide local trophic support.95,96 They are most commonly found around the central canal of the spinal cord and mobilize after injury; however, their numbers are limited making transplant of exogenous adult- or stem cell-derived NPCs a promising strategy.80,97 In animal models of cervical and thoracic SCI, transplanted NPCs have been shown to reduce cystic cavitation, remyelinate denuded axons, and improve behavioral outcomes over time.80,98 In 2016, 2 phase II trials led by StemCells Inc. (Newark, California) were terminated early due to insufficient efficacy. The studies were assessing the effects of human CNS stem cell transplants for thoracic (NCT01321333) and cervical (NCT02153876) SCI.83 The results of these trials have not yet been published; however, interim reports provide evidence that intraparenchymal cell transplants are feasible in humans. Based on emerging preclinical data, it is likely that further modifications to the transplanted cells and/or their local environment will be necessary to enhance motor outcomes. Oligodendrocyte Progenitor Cells Oligodendrocyte progenitor cells (OPCs) have similar multipotent potential to NPCs, but they preferentially differentiate to oligodendrocytes to remyelinate axons. Several preclinical studies have found remyelination and enhanced functional recovery after OPC transplantation.99,100 A phase I/II open-label trial (n = 35; NCT02302157) is now underway by Asterias Biotherapeutics (Fremont, California) to assess allogenic OPCs with results expected by 2018.83 Biomaterials Technological advances have fueled the emergence of several new classes of biocompatible biomaterials with direct applications to SCI. These materials can be seeded with stem cells, engineered to deliver growth factors, and can even be made to biodegrade over time. Moreover, they are increasingly being designed to fill cavitation defects with a structure that closely mimics the native extracellular matrix.101-105 In rodent models, biomaterials such as hyaluronan-methylcellulose,102 fibrin-thrombin matrices,106 and self-assembling peptide QL6107 have all been shown to improve histological and behavioral recovery. Closer still to clinical translation, Neuro-Spinal Scaffold (NSS) is a poly-L-lysine and poly(lactic-co-glycolic acid; PLGA) polymer blend currently in phase III trial by InVivo Therapeutics (Cambridge, Massachusetts; n = 20; NCT02138110). The trial will examine the effects of NSS in individuals with AIS grade A T2-T12/L1 injuries and contains no control arm as a Humanitarian Device Exemption was provided by the FDA making this a Probably Benefit Study. The study is expected to conclude by 2017.83 Physiological Approaches Therapeutic Hypothermia Rapidly decreasing core temperature to 32 to 34°C via a combination of ice packs, cooling blankets, and/or intravascular catheter cooling has been shown to reduce CNS injury after in-hospital cardiac arrest and neonatal hypoxic-ischemic encephalopathy.108-110 These temperatures dramatically reduce the basal metabolic rate of the demanding CNS and attenuate the systemic inflammatory response.111 When applied to SCI, preclinical data suggest there is local tissue sparing and improvements in behavioral recovery in the long term. In patients with AIS grade A injuries, a pilot study (n = 14) found early therapeutic hypothermia (TH) to be associated with better neurological outcomes.112,113 A larger phase II/III trial by the Miami Project to Cure Paralysis entitled “Acute Rapid Cooling for Traumatic Injuries of the Cord”is currently planned. The study will assess varied durations of TH initiated within 6 h of injury to confirm both efficacy and treatment time. Cerebrospinal Fluid Drainage Ischemia is known to be a critical component of the secondary injury cascade. Similar to MAP elevation, cerebrospinal fluid drainage (CSFD) attempts to improve early spinal cord perfusion pressure to reduce the ischemic territory. While an initial 2009 trial (n = 22) failed to find efficacy, recent large-animal studies have found that CSF drainage and MAP augmentation can act synergistically to enhance spinal cord blood flow.114,115 A phase IIb (n = 60; NCT02495545) randomized trial combining CSFD and MAP elevation is now underway to determine if the treatment can improve 6-mo neurological outcomes for patients with acute AIS grade A, B, or C injuries from C4-8. The study is expected to conclude in 2017.83 CONCLUSION The progress of SCI research is fast and novel findings are being paired with lessons learned from unsuccessful SCI clinical trials. To achieve success in future clinical trials in SCI, we should consider the heterogeneity of patients, pathophysiology, and response to treatment. In keeping with this, strict limits relating to level of injury as well as ASIA grade have been set in some of the forthcoming clinical trials including the cethrin and riluzole trials. The strategic management of SCI going forward is likely to involve the administration of drugs to mitigate the secondary injury at the acute phase, followed by cell transplantation therapy to regenerate damaged spinal cord tissue from subacute to chronic phase.18,19 We believe that the therapeutic approaches discussed in this review and the continuous efforts in basic and clinical research are creating a new path to regenerative medicine for SCI. Disclosures This work is funded by Canadian Institutes of Health Research, AOSpine North America, Rick Hansen Institute, Halbert Chair in Neural Repair and Regeneration, and Dezwirek Foundation. The senior author (MGF) acknowledges support from the Halbert Chair in Neural Repair and Regeneration and the Dezwirek Foundation. Dr Fehlings is a consultant (speaker) for Pfizer and Zimmer Biomed and a consultant for InVivo Therapeutics. The other authors have no personal, financial, or institutional interest in any of the drugs, materials, or devices described in this article. Acknowledgements We thank Dr. Madeleine O’Higgins for copyediting this manuscript.