Whereas loss of sensory and motor function are important manifestations of nerve injury, often one of the most troublesome complications is pain. In evaluating this problem, it is well to begin with considering how nerves cause pain.
 

Etiology of Pain

 
Pain normally arises from activity in specialized nerve fibers, termed nociceptors. The question arises, does nerve injury pain represent an activation of nociceptors (an active process), or does nerve injury pain arise from loss of afferent input to the spinal cord (a release mechanism, i.e. a passive process)? Prevailing evidence favors the first mechanism in most situations. Namely, the nerve injury somehow leads to activity in nociceptive fibers and thus induces the painful condition. This has important therapeutic implications: The first, and foremost goal of therapy is to eliminate the nociceptive input arising from the nerve injury.
An important feature of nerve injury pain is hyperalgesia (hypersensitivity). In hyperalgesic states, the normal relationship between stimulus intensity and neural response is enhanced. Stimuli that ordinarily are considered nonpainful cause pain. Nociceptors in the periphery can be sensitized (i.e. develop an enhanced response to natural stimuli). This sensitization provides a basis for certain aspects of hyperalgesia due to inflammation. When nociceptors activate certain central cells, a central form of sensitization occurs such that input from receptors other than nociceptors induces pain. Thus, patients with nerve injury may have what is termed "touch evoked pain" (lightly touching the skin hurts; sometimes termed allodynia). The region wherein this mechanical hyperalgesia occurs typically includes the area innervated by the nerve but may also extend far outside this area. In some cases, cooling, as well as mechanical stimuli, induces excessive pain. As will be described subsequently, this phenomenon of cooling hyperalgesia is a clue that the sympathetic nervous system may play a role in the pain.

Pain and the Electrophysiology of Nerve Injury

 
Nerve injury pain is a positive phenomenon-there must be activity in nociceptive fibers. This can occur in one of several ways. One mechanism concerns the capacity, in some patients, of the sympathetic nervous system to induce activity in nociceptive fibers. Nerve injury can also induce neural activity by causing ectopic generation of action potentials (ectopic generators). That is, fibers of passage at the injury site can generate neural activity spontaneously. Surprisingly, electrophysiologic data from nerve fibers that end in a neuroma suggest that ectopic generators are of equivocal importance in nerve injury pain. In many studies few fibers that end in a neuroma created by nerve ligation are spontaneously active.

 
Another related mechanism is that of ectopic excitability. This refers to the capacity of the fibers to be activated by stimuli applied to the nerve trunk. Thus, when there is ectopic excitability to mechanical stimuli, fibers may be activated in the region of injury either by endogenously or exogenously generated mechanical stimuli. Electrophysiologic evidence suggests that ectopic excitability is an important mechanism of pain in nerve injury.

 
A third mechanism refers to the compelling clinical evidence that nociceptive fibers innervate the nerve trunks themselves. Thus, at a compression or injury site, the nerve is quite often tender. Pain is therefore described as originating at the compression site (rather than the area innervated by the involved nerve). This is likely due to nociceptive fibers that constitute the nervi nervorum. The nervi nervorum fibers, which innervate the epineurium of the nerve, are activated by the entrapment or by other mechanical factors that induce traction on the nerve trunk.
Yet another mechanism for pain from nerve injury involves the phenomenon of cross talk. Cross talk refers to cross activation of nerve fibers within the nerve trunk. Evidence suggests that cross talk occurs at nerve injury sites. The importance of cross talk however, as a mechanism for pain is far from clear.
 

Nerve Injury Milieu
 

As noted previously, entrapment of a nerve may activate the nervi nervorum fibers and cause pain. When this happens, the nerve is locally tender. Thus local tenderness may be a harbinger of nerve entrapment. Entrapment is likely the most common manner in which a nerve injury causes pain. Recently, an animal model of nerve injury pain was introduced, whereby ligatures are placed loosely around the sciatic nerve in rats. The animals display behavioral signs of hyperalgesia similar to what is observed in humans. This model will provide investigators a potent means to investigate further the mechanisms for pain, both peripherally and centrally. In addition, there is great promise that this model will lead to more effective treatments for neuropathic pain."
When a nerve is entrapped, axonal transport is blocked. This leads to the so-called Tinel's sign. Tinel's sign results when the nerve is ectopically sensitive (ectopic excitability). When the skin is tapped briskly over the entrapment site, the patient reports paresthesias in the area served by the sensory fibers in the nerve. Tinel's sign combined with local tenderness constitutes strong evidence for nerve entrapment, irrespective of other data. Of course, in many instances nerve entrapment leads to neurologic impairment. Sensory loss can be assessed along with motor impairment. Sensory deficit, atrophy, and muscle weakness, therefore, can also serve as indicators of nerve entrapment.
Electrophysiologic studies supplement the diagnosis of nerve entrapment syndromes. Slowed conduction often can be demonstrated at the point of entrapment. In advanced cases the electromyographic study will reveal denervation in the relevant muscles. A negative electromyogram and nerve conduction study do not exclude the diagnosis of nerve entrapment. 
 

Structural Nerve Injuries

Neuromas

When a nerve is severed and the proximal axons are still in continuity with the dorsal root ganglion, neuromas will form. A neuroma is a densely packed cluster of regenerative sprouts that arise when there is discontinuity of the nerve, usually secondary to trauma. In addition, fibrous tissues proliferate at the point of nerve injury. The fibrous tissue may contribute to the blockage of successful regeneration.
Some neuromas are painful and others are not. Why is this? The key to understanding this inconsistency emerges from consideration of the electrophysiologic data noted previously. Ectopic excitability rather than ectopic spontaneous generation of action potentials emerges as the dominant finding in electrophysiologic recordings from traumatically injured nerves. The clinical corollary is that the milieu of a neuroma rather than the neuroma itself determines whether pain develops. Thus, in many, and perhaps the majority of cases, pain results due to the location of the nerve. Neuromas tend to be painful in locations where the severed nerves are subjected to excessive mechanical forces. This is particularly the case with nerve injuries in the hand and foot.
Many surgical techniques that are used to treat neuromas have been described. Simply excising the neuroma does not permanently eliminate the neuroma. As long as the nerve fiber has a viable cell body in the dorsal root ganglion and distal regeneration does not occur, a neuroma will form. The goal of surgical therapy, in these cases, is to relocate the neuroma to a different site. This site should be away from scar tissue, moving structures such as tendons, and joints. In addition, the neuroma bed should be protected from external stimuli. These objectives can be achieved in a number of ways. One is to put the severed nerve into deep muscle that has limited excursion. In certain instances it may be worthwhile to consider a nerve graft repair, if only to prevent the neuroma from reforming.

Painful Nerve Lesions in Continuity

In certain instances the nerve is injured but still in continuity. Perhaps some of the axons are severed, but the nerve is still functional. As with neuromas, pain may or may not be a problem. Where pain is present, the external milieu of the nerve may be responsible. In other cases the internal matrix of scar in the nerve may somehow promote activity in nociceptive fibers. The relative role of these two mechanisms may be difficult to discern on clinical grounds.
The first issue to consider is whether the nerve is entrapped. The nerve may be bound to adjacent tendons and scar, such that movement of the extremity results in the application of shearing forces to the nerve. Muscle flaps may be rotated to protect the nerve. This technique has been applied most frequently with the median nerve at the wrist, where the abductor digiti muscle may be rotated over the nerve. The muscle may provide a useful cushion and may help keep the nerve from becoming stuck to the overlying skin. The long-term role of this procedure in relieving pain, however, is unsubstantiated.
In some cases nerve graft repair has been successful. In these cases the nerve is not entrapped and yet pain is severe. There have been several cases where the author has grafted the injured nerve to achieve pain control, even when the nerve had some function. Grafting may work by lengthening the nerve (thus eliminating tension), by removing the scarred focus, or by facilitating successful regeneration.
Neurotomy (proximal nerve resection) may be the treatment of choice in many circumstances. For example, during inguinal herniorrhaphy, the ilioinguinal or genitofemoral nerves or both may be injured. Whether the nerves are directly injured, entrapped, or severed is seldom clear. These patients have persistent pain in the inguinal area that often spreads to the labia or testicle with or without a sensory deficit. If the pain is temporarily abated with anesthetic blocks of the ilioinguinal and genitofemoral nerves, the ilioinguinal nerve is severed near the anterior superior iliac spine, and at the same session, the genitofemoral nerve may be severed in the retroperitoneal space as it courses longitudinally on the psoas muscle.
Another example is an injury to the palmar cutaneous nerve. This nerve may be injured in the course of carpal tunnel surgery. Patients with this injury present with pain over the palmar surface, perhaps with profound hyperalgesia, with or without an obvious sensory deficit. Often there will be tenderness at the wrist crease. The problem can be solved with resection of the palmar cutaneous nerve at its origin from the median nerve 6-7 cm proximal to the wrist crease.
The principle involved in each of these examples is the same: Find a different milieu for the nerve; one free of tension, scar, tethering, motion, and excessive external mechanical stimuli. Neurotomy is particularly applicable when the nerve involved does not serve an important function. With major nerves this must be done only after extended deliberation with the patient. When a neurotomy is performed, a proximal nerve block (regional anesthesia) may help prevent pain from recurring (see discussion of phantom pain).

Palliative Measures for Controlling Nerve Injury Pain

Antidepressants

In some cases of nerve injury, despite a diligent effort, the pain cannot be solved by direct surgery. In these cases pharmacologic management may be indicated. Several drugs are worthy of consideration. Tricyclic antidepressants are of proven, though modest, value in this regard. Amitriptyline is most frequently used, though sedation can be an unacceptable side effect. Nortriptyline is much less sedating and appears to offer comparable results.
 

Anticonvulsants

Initial experience with clonazepam suggests that this benzodiazepine agonist, used often as an anticonvulsant drug, also may be of value. Sedation again is a limiting side effect. Occasionally other anticonvulsants, such as carbamazepine, neurontine and epanutin are effective. Some believe that clonazepam is particularly useful for shooting or stabbing pain.
 

Narcotics
 

Narcotics have a role in the treatment of nerve injury pain. The approach is similar to that used for treating cancer pain. The goal is to maximize quality of life. Clearly there will be a trade-off between side effects and pain relief. Striking a balance is sometimes a daunting task. Patients with a history of substance abuse probably should not be given trials of narcotic therapy. Short-acting narcotics (e.g. oxycodone) have a limited role. One should use drugs, such as morphine or methadone, which have a relatively long serum half-life. The intrathecal delivery of narcotics underwent evaluation in many centers. Too little experience exists to draw conclusions regarding longterm efficacy of this technique of drug delivery. 
 

Electrical Stimulation
 

Electrical stimulation also plays a role in the treatment of chronic pain from nerve injury. In some cases the stimulation can be applied directly to the involved nerve (usually proximal to the lesion), and in other cases spinal cord stimulation is technically more appropriate. In either case, a trial of stimulation before implantation is appropriate. A presurgical trial of transcutaneous electrical stimulation is also appropriate. Failure of this modality, however, does not preclude the success of an implanted system.

DREZ Operation for Brachial Plexus Avulsion

Avulsion injury of the brachial plexus deserves special mention. Avulsion injuries involve a severe stretch of the brachial plexus, such that the roots are torn from the spinal cord. Three types of brachial plexus avulsion injuries have been defined: upper plexus avulsion, lower plexus avulsion, and avulsion of the entire plexus. Severe, disabling pain frequently results. The mechanism is not understood; however, one compellingly simple explanation is that an epileptic focus occurs in the injured dorsal horn. Since one of the primary functions of the dorsal horn relates to pain sensation, the outcome of this epileptic discharge is pain.
In these cases the elimination of the epileptic focus is an appropriate goal of therapy. This can be accomplished via the so-called DREZ (dorsal root entry zone) operation. A series of microlesions are placed in the dorsal horn area from where the roots were avulsed. Pain relief was observed in 85 % of the cases. The operation, however, is a serious undertaking. It entails risks of injury to the subjacent corticospinal tract.

DREZ Operation for Nonbrachial Plexus Avulsion
 

The success of the DREZ operation for brachial plexus avulsion has encouraged trials for other conditions. Some successes have been reported with postherpetic neuralgia, but the morbidity associated with treatment, particularly in older patients, is high. In addition, efficacy appears to be substantially less than that associated with similar surgery for post brachial plexus avulsion pain. In paraplegics the DREZ operation seems helpful in cases where patients have particular problems with radicular pain referable to the level of spinal cord injury.
The role of the DREZ operation to treat pain from lesions of nerves distal to the dorsal root ganglion is unclear. It is not observed that the DREZ operation to be associated with clear success in this patient population. Furthermore, for reasons that are unclear, some patients seem (at least temporarily) to have worse pain after a DREZ operation.

Prevention and Treatment of Stump Pain and Phantom Pain
 

The origin of the pain in some cases of nerve injury is central. Anesthetic blockade of the injured nerves in these cases does not relieve the pain. The origin of this pain is based on neural events that are generated within the central nervous system (CNS) and that are independent, at least to some extent, of peripheral nervous system input. Such a scenario occurs in the context of phantom pain, with or without accompanying stump pain. Stump pain by itself sometimes can be helped by proximal neurotomy, though at a price, if the stump is rendered insensible. Proximal neurectomy is not generally helpful in the case of phantom pain.
The presence of pain at the time of neurotomy (done as part of the limb amputation) increases the chance that phantom pain will occur. If the amputation is performed under regional anesthesia, the occurrence of phantom pain appears to be less prevalent. It is as if to suggest that noxious inputs into the CNS establish a sustained CNS pain engram that persists if a neurotomy is done in this circumstance. This suggests that it may well be prudent to block peripheral nervous system input to the CNS at the time of an operation that entails a peripheral nerve section, particularly if that nerve is the source of pain.
 

Regeneration Pain
 

Sometimes quite severe pain can arise from a nerve injury in the context of nerve regeneration. Presumably the nociceptive fibers in the process of regeneration in some patients become ectopically active. The problem is self-limited. As the regeneration is completed, the pain abates. It is extremely important to recognize the potential for this occurrence. Surgery is obviously most inappropriate and can threaten the regenerative process. The clinician must assess whether the pain is occurring in the context of a nerve lesion that is getting better or in the context of a nerve lesion where regeneration is not proceeding satisfactorily.
 

Pain and the Sympathetic Nervous System
 

In all cases where pain is associated with nerve injury (and other situations as well), one must consider whether the pain is sympathetically maintained (sympathetically maintained pain [SMP]). Patients with SMP have pain that is dependent on sympathetic efferent function. This disorder is sometimes referred to as reflex sympathetic dystrophy or causalgia. These terms are avoided here because the linkage to the function of the sympathetic nervous system with these designations is not clear.

Mechanism of SMP

We are very close to understanding SMP at the molecular level. Several observations are noteworthy: (1) Stimulation of the peripheral, but not central, cut end of the sympathetic chain reproduces pain in SMP patients after sympathectomy. (2) Local anesthetic blockade of the appropriate sympathetic ganglia (by definition) rapidly abolishes SMP (a series of blocks may do so permanently). (3) Depletion of peripheral norepinephrine by regional intravenous guanethidine abolishes SMP (again, a series of such blocks may achieve sustained relief). (4) Intradermal injection of norepinephrine rekindles the pain and hyperalgesia previously relieved by a sympathetic block. Norepinephrine does not cause pain in normal individuals. (5) Anecdotal reports suggest that alpha-adrenergic antagonists, including phenoxybenzamine and the alpha₁-adrenergic antagonist, prazosin, may be effective in relieving pain in patients with SMP (6) The skin in the area affected by pain may be abnormally cold in patients with SMP. Cold skin in a patient with pain, however, does not indicate in and of itself that the patient has SMP Signs of increased sympathetic discharge do not provide a sufficient basis establishing the diagnosis of SMP.
Taken collectively, these observations suggest that SMP is an alpha-adrenergic receptor disease. Moreover, activation of these adrenergic receptors in patients with SMP evokes pain via activation of nociceptors. It may be that the adrenergic receptors are expressed directly on the nociceptors themselves. One conjecture is that nerve injury evokes a genetic signal in the dorsal root ganglion (as a consequence of neural activity), such that alpha receptors are sensitized and transferred to the peripheral terminals of the nociceptive fibers, i.e. the signal for the production of alpha-adrenergic receptors in nociceptive afferents is a nerve injury. It is hypothesized that injury produces neural activity in the nociceptive fibers, and this is a genetic signal in the dorsal root ganglion cell such that alpha-receptors are synthesized in abundance. After synthesis, the receptors are transported to the peripheral terminals of the nociceptive fibers and they thus make the nociceptors vulnerable to excitation via the local release of norepinephrine. The underlying disease heals. That is, the initial injury that incited the activity in the nociceptors goes away. The nociceptors remain active, however, as a consequence of the sympathetic activity in the region of the nociceptors. Further activity in the nociceptors serves as an ongoing genetic signal providing a further impetus for production of alpha-adrenergic receptors. Thus, a vicious cycle results.
This scenario helps explain the remarkable finding that one or a series of sympathetic blocks sometimes leads to a sustained relief of SMP By blocking the activation of the alpha-receptors via a sympathetic block or other sympatholytic therapy, the excitation of the nociceptors is eliminated, and the genetic signal conveyed via the neural activity in the nociceptive fibers in the dorsal root ganglion cells is eliminated.
A key component of this hypothesis is that alpha-receptors in the skin and in adjacent tissues become linked to the activation of nociceptors. Alpha receptors are of two types: the alpha₁ and alpha₂ receptors. It was determined that the topical application of the alpha₂ -adrenergic agonist, clonidine, eliminates the hyperalgesia in the region of the application. It is reasoned that clonidine relieves pain in the vicinity of the application by locally blocking the release of norepinephrine via activation of alpha₂ receptors located on the terminals of the sympathetic efferent fibers. The administration of the alpha₁-selective agonist, phenylephrine, in the clonidine-treated area rekindled the pain previously extinguished by the topical clonidine. This evidence, therefore, supports the concept that SMP is an alpha₁-adrenergic disease. Topical treatment with clonidine in the hyperalgesia zone may be a useful treatment in some patients, in particular when the area of pain is very small.
 

Diagnosis of SMP
 

The ideal test in medicine is characterized by the following features: sensitivity, specificity, and safety. The traditional techniques for diagnosis of SMP are the local anesthetic sympathetic ganglion block (LAB) and the regional infusion of guanethidine (RIG). Both of these tests have problems with regard to sensitivity, specificity, and safety.
The LAB may fail to block properly the ganglion. When fluoroscopy is used as an adjunct to the performance of the LAB, the likelihood of missing the ganglion with the anesthetic is lessened; however, problems may still arise with respect to target localization. It is well to recall that in performing a stellate ganglion block, the desired target is the T2 ganglion, since the T2 ganglion supplies the sympathetic innervation to the hand.
Specificity is a problem with both the LAB and the RIG. No mechanism exists for interpreting the placebo response with either test. The existence of the placebo response is a most important point to consider when interpreting a test that concerns pain relief. Also, with the LAB procedure it is well to keep in mind that lidocaine and its analogues may attenuate, via a systemic effect, neuropathic pain. Similarly, lidocaine is usually given with the RIG in order to decrease pain from the initial norepinephrine release. This compromises interpretation of the test. Likewise, the tourniquet inflation applied during the RIG blocks conduction in sensory fibers and may, in and of itself, affect pain. An important and frequent problem with LAB is that the somatic roots that serve the painful area are very near the sympathetic chain. It may be difficult to determine whether the pain relief that is achieved from LAB is from the sympathetic block or from somatic blockade.
Although generally safe, both the RIG and LAB have been associated with several complications. These include injury to the recurrent laryngeal nerve, pneumothorax, inadvertent vascular injection, puncture of the kidney, and leakage of guanethidine into the systemic circulation. In addition, both tests are considered unpleasant and frequently are poorly tolerated by the patient.
 

Alternative Test for SMP:
 

Systemic Infusion of Phentolamine
 

If SMP is an alpha,-adrenergic receptor disease, then a logical test for the diagnosis of this disorder would be to administer a short-acting alpha-adrenergic blocking drug such as phentolamine. Patients with SMP should derive at least short-term benefit from an intravenous infusion of phentolamine.
To study this, the effects of systemic intravenous phentolamine with a local anesthetic ganglion block in patients suspected of having SMP were compared. The pain relief obtained from the two tests was highly correlated.
Phentolamine infusion has a number of advantages over LAB: (1) It is essentially a painless test. It requires only an intravenous line through which the medications are given. (2) It is conducive to placebo controls. The patient has no knowledge as to when the phentolamine is administered. Thus, the effects of saline infusion can be tested prior to or following the administration of phentolamine. (3) It is likely to be more specific than other tests in that it targets specifically the alpha receptor. (4) To date no complications have occurred with the phentolamine test.
Tachycardia is a prominent effect of phentolamine administration. This can be avoided by simultaneously administering propranolol (which appears to have no effect on the pain relief evoked by phentolamine). The protocol for administration of phentolamine is listed in Table 1.

Cold Hyperalgesia Test
 

Remarkably, all patients with SMP due to nerve injury appear to have exquisite pain following the application of mild cooling stimuli (cold hyperalgesia). This can be demonstrated in a number of ways. First, patients will usually offer as their own observation that cooling greatly exacerbates the pain. Second, patients can be shown to have cold hyperalgesia by simply comparing the effects of running cold and warm water from the tap onto the affected area. Third, one can examine the effects of placing volatile substances such as acetone on the skin. If the patient does not have cold hyperalgesia, a search for SMP will not be fruitful. It is to be noted, however, that many patients who have cold hyperalgesia do not have SMP; therefore, though the test for cold hyperalgesia is a sensitive test for SMP, it is by no means specific for SMP
 

Management of SMP
 

Once the diagnosis of SMP is established, several treatment options are available. Since the culprit is the alpha-receptor, a major strategy of therapy is to block the activation of this receptor. This strategy has been termed sympatholysis.
In many situations SMP may be eliminated permanently by instituting a sustained period of sympathetic block. The length of that critical period varies from patient to patient. In other patients the duration of pain relief never outlives the interval of sympathetic blockade. Several interventional approaches that will achieve temporary sympatholysis are possible. An epidural catheter may be placed and the patient may have injection of local anesthetic through the epidural catheter for a period of hours or days. Alternately, a series of local anesthetic ganglion blocks can be performed over several days or weeks. A series of regional blocks with drugs such as guanethidine (that deplete the sympathetic terminals of norepinephrine) also may be done.
Sympathetic efferents course with the major nerves; therefore, a sympathetic block can be achieved merely by performing a nerve block. If the patient's pain is located in the distribution of the median nerve, it might be most appropriate to simply perform a median nerve block just proximal to the painful area. It is well to remember that even if the nerve block is applied distal to the locus of nerve injury, pain relief can be obtained in patients with SMP.

Oral sympatholytic therapy may be useful in certain patients. The target, once again, is the alpha-receptor. Oral phenoxybenzamine treatment blocks both types of alpha receptors. Prazosin more specifically targets the alpha-receptor. Oral clonidine theoretically may be of some benefit. The problem with each of these drugs is that side effects may be troublesome and prohibit achieving a dosage that is adequate to block the target receptors. This is certainly not the case in all patients. Oral therapy is especially appropriate if one can use the treatment for a limited period of time, say a few weeks, and then discontinue the therapy with the acquisition of a sustained relief of the pain. Oral therapy, as a long-term treatment, is usually an untenable alternative due to side effects.

Surgical sympathectomy is an excellent treatment for SMP in situations where sympatholysis fails to achieve long-term benefit. It has been the author's experience that SMP can be eliminated with a properly performed sympathectomy. The "bad name" that surgical sympathectomy has acquired in some circles is undoubtedly due to inappropriate patient selection (a very common problem) or (in some cases) to an inadequately performed sympathectomy.

In many cases where the SMP affects the lower extremity, it is necessary to do a bilateral lumbar sympathectomy. A typical scenario is that of a patient who may have been successfully relieved of his or her pain for several weeks after an ipsilateral sympathectomy only to have a return of pain. In each case, performance of a contralateral lumbar sympathetic block has once again removed the pain. This has occasionally led to performance of a contralateral lumbar sympathectomy.

It has been the author's practice to excise approximately 8-9 cm of the sympathetic chain. This generally will include three ganglia. The lumbar sympathectomy is performed through a standard retroperitoneal approach via a flank incision. Patients frequently develop groin pain and pain in the proximal thigh after the operation. This usually lasts several weeks and then goes away. There has been much speculation about the cause of this proximal pain. The pain fits the distribution of the lateral femoral cutaneous, ilioinguinal, and genitofemoral nerves. It is suspected that this pain problem is a consequence of the stretching of these nerves during surgery.

The sympathectomy procedure for the hand can be performed by a variety of approaches. One such technique is to do a costotransversectomy via the posterior midline approach. The T2 ganglion is found underneath the second thoracic nerve root. Removal of the T2 ganglion generally suffices for treatment of SMP affecting the hand.

Some have advocated percutaneous techniques for the performance of a sympathectomy. Probes are placed in the region of the sympathetic ganglia. Thermal coagulation is then instituted to achieve the sympathectomy. This procedure offers the advantage of sparing the patient a major operation. In certain cases, however, this type of procedure will lead to only a temporary sympathectomy and it may be necessary to repeat the procedure.

Endoscopic T2 ganglionectomy is an alternative approach. Again, this may save the patient a major operation.
 

Limitations of Treatment

In some cases it is likely that the nerve injury-induced genetic signal is not susceptible to being turned off. In these cases the sympathetic block fails to induce a sustained period of pain relief that outlasts the duration of the pharmacologic action of the sympatholytic therapy. In addition, it is well to keep in mind that in some cases the injury does not go away. If, for example, the cause of the nociceptive discharge is an injury to the nerve (i.e. an entrapped nerve), sympatholytic therapy may eliminate some of the patient's pain, but will not eliminate that component of the pain that is due directly to the nerve injury (the nerve entrapment). Sympatholytic therapy only addresses that aspect of the pain that is linked to activation of the alpha₁-adrenergic receptors.

Sympatholytic therapy seems to give rise only to partial pain relief. Phentolamine administration, for example, may lead to only 30% to 50% pain relief on a consistent basis. This is observed similarly with the local anesthetic sympathetic ganglion blockade. If a surgical sympathectomy is ultimately performed, it is likewise anticipated that the patient will derive only partial pain relief. If, however, the clinician is directed to the underlying disease, i.e. the nerve entrapment, perhaps both the SMP and the underlying mechanism for activation of nociceptors may be eliminated.

In many cases the diagnosis of SMP explains only part of the patient's pain. The clinician needs to keep in mind that there may be other sources of pain in patients with SMP. The phrase "sympathetically independent pain" may be used to refer to that aspect of the pain that is not maintained by sympathetic function.

Summary and Conclusions

One of the major challenges regarding chronic pain management is the establishment of the diagnosis. Nerve injury or nerve entrapment needs to be considered in the differential diagnosis when a patient has chronic and otherwise enigmatic pain. Treatment of nerve injury pain requires a detailed understanding of peripheral nervous system anatomy. The role the sympathetic nervous system plays in each case must be ascertained. Most patients can be helped if the clinician pays careful attention to detail. A multitude of therapeutic choices are available, ranging from simple disentrapment operations to spinal cord stimulation with implanted electrodes. In some cases, nerve reconstruction operations are worthy of consideration. In addition, many drugs have been documented to be useful for the treatment of nerve injury pain.

Current clinical evidence points to SMP as being a disorder linked to activation of alpha₁-adrenergic receptors. The alpha₁-adrenergic receptors, when activated, lead to stimulation of nociceptive fibers. Pain thus ensues. SMP can be diagnosed through systemic administration of phentolamine. This is a simple and well tolerated test that lends itself to placebo controls. Once the diagnosis of SMP is established, a number of choices are available in terms of treatment. Often short-term sympatholytic therapy will lead to sustained pain relief. In these cases, that is all that is required. Short-term sympatholytic therapy can be achieved through epidural techniques, conventional nerve blocks, and oral sympatholytic therapy with drugs such as phenoxybenzamine and prazosin. Topical clonidine application to the painful area also appears to work in some patients. At times, surgical sympathectomy is necessary to remove the SMP on a permanent basis. In well-selected patients, the results of surgical sympathectomy are excellent. Many patients with SMP have pain that is independent of sympathetic function. It is also important to keep in mind that SMP may arise through an underlying disease such as a nerve entrapment which, if corrected, will also lead to the alleviation of the pain.