The man on fire syndrome is very rare. Neuropathic pain is not. It occurs commonly in disorders as diverse as traumatic limb amputations, nerve or nerve root compression, and peripheral neuropathy due to many causes. The central molecular player in inherited erythromelalgia, NaV1.7, is pivotal in these disorders too.
The eminent nineteenth-century physician Silas Weir Mitchell—one of the founders of the American Neurological Association—holds a place of special interest to military historians because he tended to wounded soldiers on the Civil War battlefields. At that time, amputation of the injured limb was all that was available to treat bullet wounds, and the concept of the ambulance—initially just a wagon to carry a wounded soldier away from the theater of active conflict—was just emerging. Mitchell was preoccupied with pain and is credited with publishing the first description of erythromelalgia (still called, by some, Mitchell’s disease), although the cases he described came from patients without a family history. Mitchell also had a deep interest in pain following trauma to nerves, which he described in detail in his book Injuries of Nerves and Their Consequences (Mitchell 1872). As a clinician who treated soldiers wounded on the battlefield, Mitchell was aware of the nerve injury that necessarily accompanies traumatic limb amputation and of the long-lasting pain that can ensue, not just in the retained limb stump but also in a phantom limb. The phantom limb is a construct of the nervous system that frequently replaces the lost appendage with the sensation of a retained hand or foot, not in reality there but nonetheless felt, as a part of the body that feels as if it were present, sometimes twisted, and often painful. Mitchell coined the term “phantom limb” and documented the pain that can come with it. He was also aware of the propensity of high velocity missiles—gunshot injuries—to produce chronic pain (Mitchell, Morehouse, and Kenn 1864).
Following nerve injury, the injured nerve fibers within peripheral nerves have a propensity to regenerate. If the proximal and distal parts of the injured nerve are lined up and reattached and there are no obstacles in the way, the regenerating nerve fibers may find their way to appropriate peripheral tissues and even reinnervate them, reestablishing sensation in the territory supplied by the injured nerve. If, however, there is scar tissue at the site of the injury or if the limb has been severely traumatized or amputated, regenerating nerve fibers may not be able to find their way to appropriate peripheral locations and can profusely sprout in a futile attempt at regeneration, forming a knot-like collection of blindly ending nerve fibers that do not reestablish connections with the periphery. These tangled masses of blindly ending axons, mixed with proliferating connective tissue, are called neuromas. The injured axons within experimental neuromas (in laboratory animals), and in human neuromas, are the sites of abnormal, ectopic impulse generation which contributes to neuropathic pain (Amir and Devor 1993; Burchiel 1988). It was known, even before the entire deck of human sodium channels was cloned and sequenced, that hyperexcitability within pain-signaling neuromas depends on the activity of sodium channels (Chabal, Russell, and Burchiel 1989; Devor, Wall, and Catalan 1992; Matzner and Devor 1994). And, once sodium channels were cloned and their functional attributes understood, there was evidence for a contribution of NaV1.7 to many types of acquired neuropathic pain as well as inflammatory pain in animal models (Black et al. 2004; Cummins, Sheets, and Waxman 2007; Rush, Cummins, and Waxman 2007). But, what about human neuromas?
In 2007, I began a collaboration with Lone Nikolajsen and Troels Jensen, two pain researchers, and their surgeon colleague Karsten Kroner, who were working at the Danish Pain Research Center at the University of Aarhus. Well-preserved human neuromas, and in particular human neuromas collected from patients who had been carefully characterized in clinical terms, were not easy to come by. My collaboration with the Aarhus pain researchers provided my laboratory with a source of tissue from well-studied patients. In a joint study (Black et al. 2008), we examined the expression of sodium channels within painful human neuromas from their patients, in some of whom there had been traumatic limb amputations.
Each of the patients described in this paper had a well-characterized peripheral nerve injury. These patients had been referred to the Aarhus Research Center because they had failed to respond to treatment with medications and had consented to participate in a clinical study in which the neuromas were surgically removed in an attempt to alleviate pain. The severity and pattern of pain had been carefully assessed by the Aarhus researchers in each patient. Following removal by the surgeon in the operating room, each neuroma was rapidly and carefully placed in a freezer and stored at –80°C until shipping by air express, on dry ice, from Denmark to our laboratory at Yale where it was cut into sections for viewing under the microscope. We used antibodies which bind specifically to each type of channel and fluoresce under the microscope (a technique called “immunocytochemistry”), to assess the presence of various types of sodium channels within axons in these precious human neuromas. As shown in Black et al. (2008), this allowed us to demonstrate that higher-than-normal levels of NaV1.7 and NaV1.8 accumulate within human painful neuromas. The presence of high levels of these channels would be expected to produce hyperexcitability in injured axons. A few years later we showed, using computer simulations and dynamic clamp, that even small elevations in the level of expression of either of these channels can produce hyperexcitability (Choi and Waxman 2011; Vasylyev et al. 2014). Interestingly, we also observed increased levels of an enzyme called MAP kinase ERK1/2 within the injured axons within neuromas. This may magnify the functional consequences of the accumulation of NaV1.7 within neuromas, since ERK1/2 is known to interact with NaV1.7 and make it easier to activate (Stamboulian et al. 2010). Irrespective of whether MAP kinase amplifies the effect of NaV1.7, our observations pointed to a molecular target—NaV1.7—in painful human neuromas.
These observations may help to explain the phantom limb phenomenon. One theory attributes the phantom limb experience to functional reorganization of the cerebral cortex, with the cortical areas originally receiving input from the amputated body part now deprived of input from that body part, and responding to other inputs to produce a message that is interpreted as indicating the presence of the lost limb (Ramachandran and Blakeslee 1998). Other experiments challenge this concept and posit a peripheral source of the abnormal pain signaling. For example, Vaso et al. (2014) provided data supporting the hypothesis that phantom limb pain is a result of exaggerated input to the brain, generated ectopically in the injured nerve fibers that, prior to injury, innervated the limb. In a series of amputees, they applied the local anesthetic lidocaine—a sodium channel blocker—to the DRG neurons whose axons had been amputated, and rapidly and reversibly extinguished phantom limb pain as well as nonpainful phantom limb sensations. The most parsimonious explanation of these observations is that, following injury to the distal ends of peripheral nerve fibers that accompanies traumatic limb amputation, sodium channels within the injured nerve cells produce aberrant barrages of nerve impulses, thereby producing pain. The importance of these results (Vaso et al. 2014), and of our observations on abnormal accumulations of NaV1.7 and NaV1.8 within neuromas (Black et al. 2008), is that they identify sodium channels as molecular targets, whose silencing would be expected to alleviate pain after traumatic nerve injury. Gene therapy approaches which “knock down” the genes for sodium channels have, in fact, been shown to alleviate pain after experimental nerve injury in rats (Samad et al. 2013). Identification of abnormal accumulations of NaV1.7 and NaV1.8 within human neuromas encourages us to think that targeting of these channels may provide pain relief not only in animal models of pain, but also in people who have sustained nerve injuries.
When Mitchell tended to wounded soldiers on the Civil War battlefield, amputation of an injured limb was one of the few options available to the doctor. Surgical excision of neuromas continues to be used for some patients today in attempts to alleviate pain. Previous assessments of the effects of surgical removal of neuromas, however, provided a mixed picture. There were some suggestions that this operation might relieve pain (Ducic et al. 2008; Krishnan, Pinzer, and Schackert 2005), but the results differed substantially in different series of patients. The patients we studied had continued to suffer from intractable and severe pain from their neuromas despite attempts at treatment with multiple medications. So, as a corollary to our study on sodium channels within neuromas, our Danish colleagues in Denmark wanted to assess the clinical efficacy of surgical excision of neuromas (Nikolajsen et al. 2010). These patients were told that surgical removal of neuromas was an experimental procedure that might result in reduced pain, but that there was a risk of worsened pain. Following surgical removal of the neuromas by a specialist in nerve surgery, patients were followed for six months, during which their pain was periodically assessed. We attempted to identify clinical features or responses to medications that might pinpoint patients who were more likely to respond favorably to surgery, but we did not find any predictors of successful outcome. We found that surgery produced long-lasting relief of spontaneous pain in only two out of the six patients we studied. Interestingly, one patient with relief of pain after surgery had a prior poor response to neuroma removal. We interpreted our findings as suggesting that, as a therapeutic maneuver, surgical excision of neuromas should be reserved for patients with intractable pain who have failed to respond to other therapies. We also noted, however, that a prior poor response to surgical neuroma removal does not preclude relief of pain after a new excision (Nikolajsen et al. 2010).
Overall, the limited efficacy of surgical removal of neuromas underscored the conclusions of our immunocytochemical study, which showed that nerve injury and neuroma formation trigger a change in the pattern of gene expression in the injured nerve cells that leads to hyperexcitability. These observations suggest that molecular targeting may be more effective than surgery in relieving pain from nerve injury and neuromas. New compounds that selectively block NaV1.7 are being assessed in clinical studies. Blockers of NaV1.8, which works in tandem with NaV1.7 to produce high-frequency firing, may come next. It would be ironic if a cure for pain, after penetrating trauma from missiles or other weapons, were to be provided not by a surgeon’s knife, but by therapeutic targeting of tiny molecules within nerve cells. That would signal a transformation of medicine from scalpel to the submicroscopic. And that seems to be where we are heading.