The use of brief post‐surgical low frequency electrical stimulation to enhance nerve regeneration in clinical practice

Abstract

Despite efforts to enhance peripheral nerve regeneration, there has been little progress in improving clinical outcomes. Recently, a method of brief post‐surgical low frequency electrical stimulation of surgically repaired nerves has been developed. It was shown to accelerate axon outgrowth across the repair site and it hastened target reinnervation. In this brief review, we describe the mechanistic insights and functional impacts of the post‐surgical electrical stimulation that have been gained through animal studies. Brain‐derived neurotrophic factor, cyclic AMP and regeneration‐associated genes play a vital role in expediting the outgrowth of axons across the injury site. The method of stimulation has also been shown to be effective in patients with severe compressive neuropathy as well as those with digital nerve laceration. Its clinical feasibility and positive impact open the door of further clinical translation in other peripheral nerve injuries.

Introduction

Despite the potential for injured peripheral nerves to regrow their axons, recovery of function following severe nerve injury in humans remains poor. In comparison with the commonly used animal models of mice and rats, the challenge of longer distances that the regenerating axons have to traverse to reach denervated targets in humans is more substantial. This is further compounded by the slow rate of nerve regeneration of 1 mm per day and a progressive loss of regenerative capacity over time and distance (Sunderland, 1947; Fu & Gordon, 1995 a,b). Therefore, despite technical improvements in surgical repair of nerve injuries and advancement in our understanding of the biology of axon regeneration, clinical outcomes have not improved substantially over the past few decades (Lundborg, 2000). The situation is more complex for transection injuries of larger nerve trunks. It is further compromised by misdirection of regenerating axons, frequently to inappropriate innervation targets that can result in incongruous movements including synkinesis, abnormal or loss of sensations (Thomas et al. 1987; Lundborg, 2000).

Following nerve injury, the denervated Schwann cells in the distal nerve stump undergo mitosis and transform from a myelinating into a growth mode by responding to neuregulin and other axon‐derived components, dividing and guiding the regenerating axons (Salzar et al. 1980; Hall, 1999; Birchmeier & Nave, 2008). However, neither the growth response of the axotomized neurons nor the Schwann cell support are sustained. Expression of the regeneration‐associated genes in the neurons and Schwann cells declines with time (You et al. 1997; Al‐Majed et al. 2000 a,b; Chen et al. 2005; Deshpande et al. 2006; Hoke et al. 2006). This explains the progressive deterioration of regenerative success of chronically injured neurons over time (Fu & Gordon, 1995 a,b; Gordon et al. 2011).

In addition to the slow rate of nerve regeneration of 1–3 mm per day, the axons that grow out from the proximal stump of a transected nerve ‘stagger’ across the site of coaptation of the nerve stumps. Consequently, the time taken for the axons to progress over distance in the distal nerve stump turns out to be much longer than predicted from the latent period of growth across the coaptation and the regeneration rate (Al Majed et al. 2000 b). Using transgenic mice whose motoneurons express yellow fluorescent protein, Witzel et al. (2005) visualized fluorescent regenerating axons as they traversed the suture site to demonstrate the ‘staggering’ of regenerating axons across the surgical gap into the distal nerve stump directly. In a rat model of nerve repair, this process was found to be surprisingly long, taking up to a month before all the regenerating axons crossed the injury site (Brushart et al. 2002). Predating these findings, Cajal illustrated this phenomenon using silver staining to demonstrate that regenerating fibres ‘wander’ across the suture site (Cajal, 1928).

Use of low frequency electrical stimulation to enhance nerve regeneration

Despite a long history of the clinical use of electrical stimulation to sustain denervated muscle bulk and to circumvent the complications of joint contractures after nerve injuries, scientific investigations of the effects of electrical stimulation of the proximal stump on axon regeneration were relatively scarce. In fact, until the new millennium, only a few studies pursued the question of a role of electrical stimulation of the proximal stump of injured nerves in promoting axon regeneration (Al‐Majed et al. 2000 a; Brushart et al. 2002) and/or target reinnervation (Nix & Hopf, 1983; Pockett & Gavin, 1985; Al‐Majed et al. 2000 a; Brushart et al. 2002; Eberhardt et al. 2006). In order to determine whether electrical stimulation accelerates axon outgrowth and/or rate of axon regeneration after nerve section and repair, we undertook a series of experiments to evaluate the effects of low frequency (20 Hz) electrical stimulation on motor and sensory axon regeneration. Using a rat model of femoral nerve transection and microsurgical repair, retrogradely transported dyes were applied to the branches of the femoral nerve in order to count the number of motoneurons that regenerated their axons across the suture site into the distal nerve stump (Fig. 1 A). Electrical stimulation of the surgically repaired nerves accelerated axon outgrowth: all the motoneurons regenerated their axons into the nerve branches within 3 weeks as compared with 8–10 weeks in the controls (Al‐Majed et al. 2000 a) (see Fig. 1 B). In the same study, we ascertained that the conduction of action potentials was essential, the effect being blocked when tetrodotoxin was applied to block conduction from the stimulating electrodes situated proximal to the nerve repair (Al‐Majed et al. 2000 a). Moreover, the effect was independent of axonal transport so that the mechanism of action of the electrical stimulation is to accelerate axon outgrowth but not to accelerate rate of regeneration of the axons once they are within the endoneurial tubes of the distal nerve stump (Brushart et al. 2002). Similar efficacy of post‐surgical electrical stimulation in mice has been reported (Ahlborn et al. 2007; English et al. 2007). Importantly, the same acceleration effects were seen with as little as 1 h of electrical stimulation for motor (Al‐Majed et al. 2000 a) and sensory nerves (Brushart et al. 2002; Geremia et al. 2007). The findings that electrical stimulation enhances nerve regeneration have since been confirmed in many animal studies as reviewed recently (Gordon & English, 2016). This opened the possibility of using electrical stimulation as a feasible intervention in clinical settings.

Figure 1. Effects of electrical stimulation on nerve regeneration in animal models and humans .

A, in a rat femoral nerve injury model, the nerve was cut above the bifurcation. Electrical stimulation at 20 Hz was applied to the proximal nerve stump for an hour immediately after suture repair. The neurons were allowed to regenerate their axons for up to 8 weeks before retrogradely transported dyes were applied to the distal stump 25 mm from the repair site. B, all retrogradely labelled motoneurons that regenerated their axons were counted in the spinal cord. Animals that received electrical stimulation showed accelerated growth taking only 3 weeks as opposed to 8 weeks for sham‐stimulated neurons in control rats. The dashed line represents mean motoneuron number in control animals. C, in a double‐blind randomized controlled trial, the same electrical stimulation treatment was applied to patients with marked median motor axon loss due to severe compression in the carpal tunnel. The extent of reinnervation was evaluated using motor unit number estimation. D, similar to the animal study data, significantly greater motor reinnervation was found in the treatment group compared with those who received sham stimulation. As in panel B, the dashed line represents mean motoneuron number in control animals. Because of the difference in size, the actual time frame of follow‐up (1 year) had to be much longer than in the rats. E, the impact of brief low frequency electrical stimulation is represented schematically. The effect of the electrical stimulation is to accelerate axon outgrowth across the repair site with the result that muscle reinnervation occurs more rapidly.

Mechanisms of action

Several lines of evidence provided mechanistic insights into how post‐surgical electrical stimulation exerts its impact on peripheral nerve regeneration. Electrical stimulation elevates neuronal expression of neurotrophic factors including brain‐derived neurotrophic factor (BDNF) and their receptors (Al‐Majed et al. 2000 a; Geremia et al. 2007). The stimulation effect on accelerated axon outgrowth can be mimicked by exogenous growth factors and, conversely, reduced by knockout of the genes for these factors in transgenic mice (English et al. 2007). Intracellularly, the elevated neurotrophic factor levels result in an upregulation of cAMP level that is likely to be mediated by calcium influx. Further downstream, this leads to sustained increased expression of regeneration‐associated genes including tubulin, actin and GAP‐43 (Udina et al. 2008). The role that cAMP plays has been confirmed by elevating cAMP pharmacologically with rolipram to selectively inhibit phosphodiesterase (Udina et al. 2010). That was found to mimic the effect of electrical stimulation.

The impacts of different stimulation frequencies on axonal growth have been examined. Udina et al. (2008) found that sensory nerve regeneration was more robust when stimulated at 20 Hz compared with 200 Hz. Indeed, high frequency stimulation sustained over a long period has been shown to be deleterious causing axonal injury possibly due to over‐activation of the Na⁺/K⁺ pump and reversal of the Na⁺/Ca²⁺ exchanger (Agnew et al. 1999; Moldovan et al. 2016).

Findings in clinical trials to date

To test the feasibility of clinical implementation, patients with severe carpal tunnel syndrome with marked motor axonal loss in the median nerve were randomized to receiving 1 h of 20 Hz electrical stimulation following carpal tunnel release surgery or surgical release alone (Fig. 1 C). Amplitude of the compound motor action potential (CMAP) was recorded on the surface of the thenar eminence in response to supramaximal stimulation of the median nerve. Amplitudes of single motor unit action potentials (MUAPs) were then recorded in response to all‐or‐none stimulation at multiple points along the course of the nerve (Fig. 1 C). Motor unit number estimation (MUNE) was done by calculating the ratio of the CMAP and the mean value of the MUAPs. The treatment group showed significantly better axonal regeneration than the thenar muscles 3 months post‐surgery. Moreover, all motoneurons regenerated and reinnervated the muscles by 12 months post‐surgery when those in the control group still had not even reached significant reinnervation even after a year (Fig. 1 C and D) (Gordon et al. 2010 a). However, despite the physiological benefits, improvement in motor performance was not significantly different between the treatment and control groups. A plausible reason is that the functional contribution of the median innervated thenar muscles to hand activity is relatively small because the majority of intrinsic hand muscles are innervated by the ulnar nerve. In addition, the forearm digit flexors are also not affected by carpal tunnel syndrome. To definitively test the effects of electrical stimulation on motor functional outcomes, other experimental peripheral injury models are needed. Nevertheless, that study served as a proof‐of‐principle that post‐surgical electrical stimulation accelerated muscle reinnervation by regenerating motor axons in humans (see Fig. 1 C and D). Although diabetes is a common comorbidity in patients with carpal tunnel syndrome, it does not necessarily compromise the effectiveness of post‐surgical electrical stimulation. Indeed, electrical stimulation has been found to enhance peripheral nerve regeneration in diabetic rats (Lin et al. 2014, 2015). A potential mechanism is through amelioration of mitochondrial dysfunction that plays a major role in the pathogenesis of diabetic neuropathy (Zenker et al. 2013). Electrical stimulation has been shown to be capable of increasing transport and biogenesis of mitochondria (Sajic et al. 2013).

In line with animal data demonstrating that injured sensory nerve fibres also respond to the 1 h period of 20 Hz stimulation with accelerated axon outgrowth (Brushart et al. 2005), a recent study was done on patients with complete digital nerve laceration who underwent early primary nerve repair. Those in the treatment group showed consistently greater improvements in all sensory modalities within 5–6 months postoperatively compared with the controls (Wong et al. 2015). Although these findings are encouraging, important questions regarding the role of post‐surgical stimulation in clinical use still remain. The following is a critical evaluation of the current state of scientific knowledge regarding post‐surgical electrical stimulation and gaps that need to be addressed.

Early versus delayed delivery of electrical stimulation

To date, all but two animal studies were done on acute nerve transection followed by immediate repair. In contrast, most major peripheral nerve injuries in humans occur in severe polytrauma settings with multi‐organ involvements and life‐threatening complications. Therefore, nerve repair is invariably delayed, often for months. Knowing that the milieu in the distal stump is fluid and its support for nerve regrowth is highly temporally dependent, a better understanding of how it could be enhanced through electrical stimulation is needed. The two animal studies that evaluated nerve regeneration after delayed repair with 1 h of 20 Hz electrical stimulation showed that electrical stimulation did have a beneficial effect in accelerating nerve regeneration (Huang et al. 2013; Elzinga et al. 2015). In humans the efficacy of electrical stimulation in promoting reinnervation of chronically denervated muscle has been demonstrated in carpal tunnel syndrome (Gordon et al. 2010 b). These findings demonstrate that the electrical stimulation paradigm can be applied to chronic models, as the symptom onset in those patients was up to 5 years.

A second important consideration is that compressive neuropathies resulting in focal demyelination and axonal loss are far more common in clinical settings. The role of electrical stimulation in ameliorating focal demyelination had not been done prior to a recent study. McLean et al. (2014) showed that electrical stimulation increased neurofilament expression and phosphorylation, which are important in axon protection. Furthermore, electrical stimulation increased expression of myelin basic protein (MBP) and promoted node of Ranvier reorganization, accompanied by enhanced clearance of ED‐1‐positive macrophages and attenuation of glial fibrillary acidic protein expression. These allowed more rapid clearance of myelin debris and return of Schwann cells to a non‐reactive myelinating state. A deeper understanding of the effects of electrical stimulation on Schwann cell behaviour may not only be important in compressive neuropathies but could also be useful in other demyelinating conditions such as Gullain–Barre syndrome.

Future in peripheral nerve repair

There are a number of potential future clinical applications that are worth investigating. Nerve reconstruction using sural nerve grafts, popularized in the 1960s, is still considered the gold standard of nerve repair. However, one of the major practical constraints is that regenerative success and functional outcomes through long nerve grafts remains poor. Since this technique involves two coaptation sites, the effects of electrical stimulation on speeding up the crossing at the suture sites could be helpful. However, this subject has not been well examined. To date, there have only been two studies by the same group that examined the role of electrical stimulation in a nerve graft model (Huang et al. 2009, 2010). While the accelerated return of motor function in the rats in which the nerve was stimulated is promising, there are important differences in the rat models used in those studies compared with patients encountered in the clinic. For example, in those rat studies, the nerve gap was bridged using nerve segments resected from the same contralateral nerve whereas clinically the use of a sensory nerve, typically the sural, is necessary. Furthermore, a gap of 10–15 mm is quite short compared with that usually encountered in the human. These are critical issues that will require more basic and clinical investigations to fully appreciate the role of electrical stimulation in nerve graft reconstruction surgery.

Distal nerve transfers have gained increasing prominence over the past decade and are now increasing in use in place of nerve graft repair. However, whether post‐surgical electrical stimulation also has a beneficial effect on distal nerve transfers has not been well established. In a recent animal study on rats, Elzinga et al. (2015) showed that electrical stimulation did have a positive effect on expediting nerve regeneration into the distal stump even when surgery was delayed. This scenario is closely akin to what is commonly encountered in clinical settings. Therefore, it will be of great interest to see whether the same benefits can be seen in patients who undergo distal nerve transfers.

Conclusions

Brief post‐surgical low frequency electrical stimulation is an adjunctive therapy for nerve regeneration whose therapeutic potential we are only beginning to fully harness. Much basic animal work has been done in the past decade to establish its mechanisms of action and to evaluate the functional impact. Although still in early stages, its potential for clinical translation has been demonstrated in two recent studies on patients with compressive neuropathy and digital nerve laceration. However, more clinical data are required before electrical stimulation will become a mainstream therapy. Although much work still lies ahead, these latter possibilities are highly enticing and could have a fundamental impact in the field of nerve regeneration research.

Additional information

Competing interests

We declare that the authors do not have any competing interest that would constitute a conflict to information provided in this article.

Author contributions

All three authors contributed to the conception and writing of this article. All authors have approved the final version of the manuscript and agree to be accountable for all aspects of the work. All persons designated as authors qualify for authorship, and all those who qualify for authorship are listed.

Funding

Canadian Institute of Health Research (CIHR) team grant: RMF82496.

Biographies

K. Ming Chan is a clinician scientist and a specialist in physical medicine and rehabilitation. His clinical focus and research interest are on peripheral nerve injury and regeneration. A medical graduate from Glasgow University in Scotland, he obtained research fellowship training in William F. Brown's laboratory at Tufts University in Boston studying neurodegenerative diseases. He heads a research laboratory at the University of Alberta and a multidisciplinary regional peripheral nerve injury programme.

Matthew Curran is a plastic surgeon resident pursuing graduate studies in K. M. Chan's laboratory investigating peripheral nerve regeneration.

Tessa Gordon is a neuroscientist trained with Gerta Vrbova at Birmingham University, England. She has done extensive basic science work on mechanistic and therapeutic aspects of peripheral nerve injury. She continues her work that began at the University of Alberta at the Sick Children's Hospital, University of Toronto. She collaborates extensively with clinicians, in particular Gregory Borschel, a plastic surgeon who heads the laboratory in Toronto, to translate promising interventions from the bench to the bedside.