Trigeminal neuralgia has afflicted human beings for centuries, and although some effective treatments have been devised for its control, its basic pathophysiology is still poorly understood. Investigators continue to study the disorder in the hope that a clearer understanding of its nature will permit the development of better forms of treatment, not only for trigeminal neuralgia but perhaps for other painful neuralgias as well.

Definition

Although facial pain of various types was mentioned in earlier medical publications, the characteristic features of trigeminal neuralgia (tic douloureux) are as follows. The patient experiences pain, usually severe, that occurs suddenly and lasts ordinarily from a few seconds to less than a minute. The pain is described as lancinating or electrical in quality, although after repetitions of such pain the sufferer may experience a more constant aching background discomfort. The pain is confined within the distribution of the trigeminal nerve on one side, it more frequently involves the lower face than the forehead and eye, and it more commonly involves the right side of the face than the left. The pain may begin spontaneously or may begin if a "trigger spot" on the skin of the face or within the mouth is stimulated by a touch, a gust of wind, a cold or hot liquid, etc. It may also be set off by speaking, chewing, or other facial movements. Trigeminal neuralgia typically recurs in paroxysms. The victim may be bothered by a series of irregularly occurring episodes of pain over days or weeks and then enjoy a pain-free period of days to months. In general. the periods of pain become more frequent with time. and pain that begins in one trigeminal division may spread to involve a larger area of the face. Bilateral trigeminal neuralgia can occur (in about 3 to 6 percent of cases). but the person usually has pain on one side for a period of time and then on the other. rather than being affected by simultaneous bilateral tic pain.

Trigeminal neuralgia is a characteristic symptom (not a disease). as just described. that affects women more than men in a ratio that has varied from 2: 1 to 4:3 in reported series. More than 70 percent of patients with trigeminal neuralgia are over 50 years of age when the disorder appears. Yet, despite its uniformity of clinical presentation. trigeminal neuralgia has been linked to a variety of etiologic agents and possible pathophysiologic mechanisms.

In most patients with trigeminal neuralgia the neurological examination is normal. However. there may be deficits that provide a clue to the specific underlying disease process giving rise to this type of facial pain. For example. the patient who has trigeminal neuralgia on the basis of multiple sclerosis may have neurological deficits resulting from involvement of other areas of the nervous system by the plaques of multiple sclerosis.

Etiology and Pathogenesis

It has long been known that tic douloureux may begin after a dental procedure in the same area of the mouth and may be misdiagnosed at first (e.g.. a "dry socket" after tooth extraction). More commonly. the pain begins spontaneously in the region of the upper or lower teeth on one side. and the patient seeks dental treatment with the assumption that the pain is of dental origin.

During the first week after the surgical treatment of tic douloureux by one technique or another. the patient may develop the lesions of herpes simplex on the face. typically about the mouth and especially on the side of the previous tic pain. This has led to the postulation that tic douloureux represents a smoldering herpetic infection of the trigeminal ganglion or peripheral trigeminal branches. The possible relation of tic douloureux to dental disease has also focused attention on the peripheral portion of the trigeminal system. as have occasional case reports of the occurrence of tic in patients with various lesions (e.g.. meningioma. epidermoid cyst. pituitary adenoma. carcinoma. aneurysm, etc.) affecting the gasserian ganglion or one or more of the trigeminal divisions. It is therefore natural that some of the initial attempts at treatment of tic douloureux involved the destruction of the "involved" branch by the injection of an agent such as ethyl alcohol or by surgical division or avulsion of the branch. However, such approaches have not shed light on the pathophysiology of the condition and have not provided a cure.

Further experience showed that in some patients with trigeminal neuralgia the trigeminal nerve is affected by a pathologic process at some point between the gasserian ganglion and the pons. For example, the trigeminal sensory root may be compressed and distorted by changes in the configuration of the base of the skull, as in Paget's disease, or by a benign neoplasm or lesion within the cerebellopontine angle (e.g., meningioma, epidermoid cyst, vestib­ular schwannoma, or arteriovenous malformation). Such benign cerebellopontine angle tumors may be found in as many as 5 to 8 percent of patients who present with tic douloureux. Frequently the tumor will give rise to symptoms and signs in addition to trigeminal neuralgia, such as reduced hearing in the ipsilateral ear or an appreciable loss of facial sensation on the same side, that provide a clue to its presence. An early age of onset of tic douloureux may signal a posterior fossa mass lesion such as an epidermoid tumor.

Surgical exploration of the cerebellopontine angle in patients with tic douloureux will frequently disclose vascular compression of the trigeminal sensory root at its entry into the pons (nerve root entry zone). It has been postulated that as people age the arteries elongate and become ectatic and the brain sags more within the skull; these two factors may bring vessels such as the superior cerebellar artery into direct contact with the trigeminal nerve. Dandy found such neurovascular relationships in 45 percent of 215 cases and postulated that vascular compression of the trigeminal root is a major cause of tic douloureux. Gardner, Jannetta, and others have pursued this idea further and have developed microvascular decompression of the trigeminal nerve as a form of treatment for tic douloureux.

Patients who are found to have fifth nerve compression by adjacent vessels usually have no neurological deficits preoperatively, although some may have a slight decrease in sensibility over the cheek. That vascular compression is not the sole cause of trigeminal neuralgia is attested to by the fact that such compression is not found at operation in perhaps 10 to 15 percent of patients who might be expected clinically to have it and that close relations between the main trigeminal sensory root at the pons and adjacent blood vessels may be found at autopsy in patients who did not have tic douloureux during life.

In addition to focusing attention on the sensory root of the trigeminal nerve as the possible key area of involvement in tic douloureux, Gardner proposed a mechanism to explain how nerve compression might give rise to paroxysmal pain. He postulated that neural compression leads to irregular demyelination within the sensory root, allowing exposed adjacent axons to come into contact, with resultant "short-circuiting" of action potentials. Such ( "cross talk" (ephaptic transmission) might be experienced as the lancinating pain of tic douloureux.

There is also evidence that tic douloureux may have its origin within the brain. For example, it has been recognized that patients with multiple sclerosis have an increased incidence of tic douloureux, probably because of demyelination along central trigeminal pathways. Approximately 2 to 3 percent of patients with tic will be found to have multiple sclerosis, and about 1 percent of patients with multiple sclerosis will develop tic douloureux. The presence of the multiple sclerosis may be suggested by an early age of onset of the tic douloureux, by bilateral tic, and by the symptoms and signs of lesions in other areas of the central nervous system besides the trigeminal pathways. Support for a central mechanism in trigeminal neuralgia can also be found in the analogy between the paroxysmal bouts of tic pain and epileptic seizures, which have a central origin, and the fact that some anticonvulsant medications such as carbamazepine and phenytoin have been found to have some effect in controlling tic pain.

It is apparent from the above that tic douloureux is a symptom that can result from any of a number of disease processes that affect the trigeminal system. The underlying pathophysiologic mechanisms remain obscure:

Any theory to explain the pathogenesis of tic douloureux must take into account the paroxysmal nature of the pain (with pain-free intervals), the facts that trigger stimuli correspond to activation of large afferent axons in the fifth nerve rather than the small nociceptive axons and that a trigger area can sometimes be in a different trigeminal division than the pain, the observation that frequently no neurological deficit can be detected, . . , and the observation that minor trauma to the region of the gasserian ganglion and/or sensory root can frequently relieve the pain. Further work is required to clarify how the various etiological factors in tic douloureux actually cause this type of pain.

Diagnosis

The diagnosis of tic douloureux is based on the history. Its typical features have already been discussed. Because of the implications regarding treatment, tic douloureux should be differentiated from other types of facial pain such as glossopharyngeal neuralgia, post­herpetic neuralgia, Raeder's syndrome, Sluder's syndrome, geniculate neuralgia, temporomandibular joint pain, cluster headaches, post-traumatic facial neuralgia, and pain due to disease of dental, orbital, or sinus origin.

The physical examination and cranial computed tomography (CT) or magnetic resonance imaging (MRI) may give a clue to the cause of the patient's pain. Usually, however, these are normal and treatment is begun.

Treatment

Pharmacotherapy

The anticonvulsants phenytoin, carbamazepine and neurontin have been found to reduce or control the pain of tic douloureux. Dilantin is less effective. The usual approach to the treatment of tic douloureux at the present time is to try one of these agents; if inadequate relief or significant side effects occur with one, the other agent is tried. Because blood levels of these drugs have not been correlated with pain relief, Loeser has recommended that the dosage of either drugs should be increased until pain relief is achieved. The standard dose of Dilantin is 300 or 400 mg/day. Tegretol should be started gradually, with an initial dose of 100 or 200 mg/day. At times, a dosage of 1200 to 1800 mg/day will be required for pain relief.

Tegretol may cause hematosuppression or hepatic dysfunction, so patients being treated with this agent should have a complete blood count and liver function studies periodically. Although Tegretol, neurontin and Dilantin are often beneficial in the initial treatment of tic douloureux, they may lose their effectiveness with time, and the patient may then require some form of surgical treatment.

Baclofen, clonazepam, and other medications have also been reported to have some value in the treatment of tic douloureux. Overall, these have not been very effective, although Lioresal may be worth trying in patients for whom Tegretol, neurontin and/or Dilantin no longer provide adequate relief. A low dosage of Lioresal (e.g., 5 mg tid) should be used at first, and this should be increased as tolerated until pain relief or a significant side effect such as excessive drowsiness is experienced. In general, analgesic medications are not effective in treating trigeminal neuralgia. They will not lessen the severe paroxysms of pain appreciably without causing significant lethargy and other undesirable side effects.

Destructive Procedures

Injections along Trigeminal Pathways

 For almost tow centuries, localized trigeminal neuralgia has been treated by the injection of alcohol into the appropriate peripheral portion of the trigeminal nerve, such as the supraorbital nerve, infraorbital nerve, second trigeminal division, or third trigeminal division. Such injections can be given quickly in an outpatient setting, they can be repeated if the tic returns, and although the instillation of alcohol is painful, this pain is transient and ordinarily is well tolerated. The results of five series reported between 1912 and 1952, involving more than 1500 patients, showed that the average duration of pain relief was, for the supra­orbital nerve, 8.5 months; for the infraorbital nerve, 12 months; for the second division, 12 months; and for the third division, 16 months. In a series reported in 1994, the median time for pain relief was 13 months for the infraorbital nerve and 19 months for the inferior alveolar nerve.

The main disadvantages of peripheral alcohol injections are the temporary sensory loss or paresthesia produced and the expected eventual recurrence of tic douloureux as the nerve regenerates and sensation returns. In addition, temporary weakness of the muscles of mastication is an expected side effect of an alcohol block of the third trigeminal division, because of the close approximation of the motor root to the mandibular nerve. Such factors have prompted Loew to state" . . . these blocks provide only temporary relief. In the hands of experienced neurosurgeons the thermocontrolled radiofrequency lesion of trigeminal ganglion or root is not more extensive surgery and is as well tolerated by elderly and medically compromised patients as alcohol injection into peripheral trigeminal branches. In the majority of cases. . . it provides permanent pain relief.,,

In an effort to achieve more permanent pain relief, various physicians have injected different types of liquids into the gasserian ganglion. Harris injected alcohol into it, starting in 1910, and subsequently others have used different destructive agents, such as hot water. phenol in glycerine. and phenol in wax. In 1940. Hams reported that he had treated more than 2500 cases of trigeminal tic by gasserian alcohol injection over a 30-year period. Among 457 patients responding to a questionnaire sent out by Hams. 316 had not experienced recurrent pain for a period of 3 years or more (3 to 31 years). By attempting complete destruction of the gasserian ganglion. Hams and others could achieve excellent pain relief with no mortality. However. the morbidity of such a procedure was substantial. with a relatively high percentage of postinjection paresthesia and pain in the anesthetic zone, a 10 to 15 percent incidence of neuroparalytic keratitis. and an expected paralysis of the muscles of mastication for about 3 months. By attempting partial gasserian destruction, several investigators found that they could reduce the incidence of these undesirable side effects. but at the price of less effective pain relief with an increased rate of pain recurrence. The gasserian ganglion is seldom injected with alcohol today. because of the development of more effective techniques with lower morbidity.

A more refined type of gasserian injection. using glycerol, was introduced by Häkanson in 1981. The development of percutaneous trigeminal glycerol rhizolysis was a classic example of serendipity. During the development of a stereotactic technique for gamma irradiation of the trigeminal ganglion and root for the treatment of trigeminal neuralgia by Häkanson and Leksell. Häkanson used glycerol as a vehicle to introduce tantalum dust into the trigeminal cistern. "The tantalum dust was used to mark permanently the trigeminal cistern for the precise stereotactic localization of the trigeminal ganglion and root. Quite unexpectedly it was observed that the intracisternal injection of glycerol alone rendered the patient completely free from the paroxysmal pain. . . . "

Trigeminal Branch Avulsion (Peripheral Neurectomy)

By dividing or avulsing a peripheral branch of the trigeminal nerve. rather than injecting it with alcohol. the surgeon can achieve a more exact. more complete. and longer lasting effect. The branches most amenable to such treatment are the supraorbital and supratrochlear/infratrochlear/lacrimal nerves, the infraorbital nerve and the inferior alveolar, lingual and mental nerves. Grantham and Segerberg reported an average pain-free period of 33.2 months after supraorbital or infraorbital nerve avulsion. The median pain-free periods reported by Quinn were. for infraorbital neurectomy. 26 months: for inferior alveolar neurectomy, 37 months: for lingual neurectomy. 38 months; and for mental neurectomy 24 months. Such neurectomies can be performed under local anesthesia in an outpatient setting. but frequently the patient is admitted to the hospital for a short stay. and sometimes a general anesthetic is used. As with alcohol injection of a peripheral branch, the main disadvantages of peripheral neurectomy are the sensory loss produced and the eventual return of tic douloureux as the nerve regenerates and sensation returns.

Retrogasserian Neurotomy, Subtemporal

For many years the standard operative approach to trigeminal neuralgia was retrogasserian neurotomy, which had its beginning in 1890. After it had been demonstrated that intracranial operations could be performed successfully, two related types of operations for tic douloureux were proposed. At first, extirpation of the gasserian ganglion was attempted. William Rose, in 1890, developed a procedure for the piecemeal avulsion of the ganglion through an enlarged foramen ovale. Because of poor exposure, frequent hemorrhage, and incomplete removal of the ganglion, this operation proved unsatisfactory. In 1891, Frank Hartley devised an extradural ­temporal approach to the gasserian ganglion to facilitate intracranial neurotomy of the second and third trigeminal divisions. This approach proved to be the technical key that opened the way for later advances. Six and a half months after Hartley's first operation, and unaware of it, Fedor Krause duplicated this operation. However, Krause carried the operation a step further in 1893 when he first completely removed the gasserian ganglion successfully. Two years later, he analyzed 51 gasserian ganglionectomies (performed by the Hartley-Krause approach) which had been reported in the medical literature, The overall mortality for these 51 cases was approximately 10 percent. Harvey Cushing then modified the Hartley-Krause approach by minimizing traction on and subsequent hemorrhage from the middle meningeal artery. The result was a reduction in mortality to 5 percent by 1905. After this, extirpation of the gasserian ganglion was abandoned in favor of the second type of operation which had been developed for tic douloureux.

Foreseeing the probable difficulties of gasserian ganglionectomy, Victor Horsley proposed retrogasserian neurotomy instead. He and William Macewen worked independently to developed such a procedure. After trial operations on monkeys and human cadavers, Horsley in 1890 attempted avulsion of the trigeminal root in a very ill woman who had had two previous extracranial operations for tic douloureux. Because of the unfortunate operative death of this patient, and the simultaneous early successes with gasserian ganglionectomy, similar attempts at dividing the trigeminal root were abandoned temporarily, David Ferrier (1890), William Spiller (1898), and Lewellys Barker (1900) each proposed that section of the root might afford a permanent cure, but it was not until 1901 that this again was attempted. In that year, Charles Frazier performed such an operation, using the Hartley-Krause approach to the nerve. His successes established retrogasserian neurotomy as the operation of choice, and later refinements minimized its morbidity and mortality.

Such refinements included the differential sectioning of the posterior sensory root fibers, the sparing of the motor root, and overall improvement in surgical operations during the first half of the twentieth century. The percentage of patients who obtained relief by subtemporal retrogasserian neurotomy ranged from 95 to 99 percent, but operative mortality remained in the 1 to 3 percent range, tic pain recurred in 5 to 20 percent, and there was significant morbidity from the procedure.

Retrogasserian Neurotomy, Suboccipital

At the same time that Frazier, Peet, and others were perfecting subtemporal retrogasserian neurotomy, Waiter Dandy developed another surgical approach to tic douloureux, the partial or total division of the main sensory root of the fifth nerve near the pons through a suboccipital craniectomy. This permitted the surgeon to spare the motor portions of the trigeminal nerve more easily. In addition, it was found that a surprising degree of facial and corneal sensation was retained postoperatively and that the incidence of neuroparalytic keratitis was reduced.

With the further refinement of surgical and anesthetic techniques and the use of the operative microscope in neurosurgical operations, Dandy's procedure has been made safer. It remains an effective way to deal with tic douloureux, especially when the neurosurgeon has exposed the fifth nerve at the pons looking for evidence of vascular compression and has not found vascular compression of the trigeminal nerve, distortion of the nerve by an adjacent tumor, or any other abnormality.

Trigeminal Tractotomy

In 1937 Sjoqvist introduced a new operation for the relief of trigeminal neuralgia, trigeminal tractotomy. This involved the surgical division of the descending tract of the fifth nerve in the medulla oblongata. It was designed to relieve pain while preserving touch sensation and trigeminal motor function. However, despite subsequent refinements, including the introduction of a stereotactic technique for producing the lesion, this procedure has never achieved widespread use, primarily because of the difficulty in dividing all of the descending trigeminocephalic tract and yet not injuring important adjacent portions of the medulla (with the unwanted production of neurological deficits such as analgesia in the distributions of the ninth and tenth cranial nerves and second cervical nerve, loss of pain and temperature sensation on the opposite side of the body, and ipsilateral proprioceptive loss and ataxia).

Percutaneous Trigeminal Radiofrequency Thermocoagulation

In the 1930s, Kirschner in Germany developed a technique for the percutaneous electrocoagulation of the gasserian ganglion. Various other physicians have since modified and perfected this approach to permit partial, precise destruction within the gasserian ganglion and sensory root. This method avoids a general anesthetic and an open operation. This technique is now in widespread use, It is especially valuable (if medication is not effective) in treating elderly persons, patients in poor general health, and patients with multiple sclerosis.

Radiosurgery - Gamma-Knife

Since its introduction by Leksell, stereotactically focused high­energy radiation has been used in a relatively small number of patients with trigeminal neuralgia to partially injure the trigeminal ganglion or sensory root. Not enough is yet known about long­term results to permit comment on the efficacy and complications of this form of treatment.

Nondestructive Procedures

All the procedures just discussed, from peripheral alcohol injections to percutaneous trigeminal thermocoagulation, have in common the destruction of some portion of the trigeminal sensory pathways. To differing degrees, they share the undesirable side effects and possible complications of such destruction, which include loss of sensation, paresthesia and pain in the denervated area (the most severe form of which is anesthesia dolorosa), and the occasional appearance of herpes simplex lesions on the face or within the mouth or nose during the initial postoperative period. For these reasons, investigators over the years have tried to develop nondestructive procedures for the successful treatment of trigeminal neuralgia.

Decompression/ Compression Operations

In 1952, Pudenz and Shelden reported the decompression of peripheral trigeminal branches at the foramen ovale or foramen rotundum in 10 patients, and Taarnhoj reported the decompression of the gasserian ganglion and posterior trigeminal root via a subtemporal approach in 10 patients. Taarnhoj's subsequent experience showed a 40 percent rate of recurrence of tic pain. In 1955, Shelden and his associates began to compress or rub the ganglion and posterior root, in the hope that such mild trauma might result in pain relief without significant sensory impairment. Further experience by various surgeons with subtemporal decompression/compression operations did show a reduced incidence of sensory loss, keratitis, and paresthesia. However, 30 to 40 percent of patients experienced some return of tic pain (usually within the first 2 years), and among 811 patients reviewed by White and Sweet, 184 (23 percent) had recurrences severe enough to require further surgical treatment.

In 1978, Mullan began to treat trigeminal neuralgia by compressing the trigeminal ganglion with a percutaneously inserted Fogarty catheter balloon that was inflated through the foramen ovale for a short period. Twelve years later, Lichtor and Mullan reported the outcome of 100 patients who were treated in this fashion and followed for 1 to 10 years. Relief persisted at 5 years in 80 percent; 4 percent of the patients reported dysesthesia; virtually all patients experienced ipsilateral trigeminal motor weakness but this resolved within 3 months. This simple technique, which is performed under a general anesthetic, now is being used more frequently while trigeminal glycerol rhizolysis is gradually being employed less often.

Taarnhoj performed decompression of the trigeminal sensory root via a suboccipital craniectomy in 20 patients between 1951 and 1959 and found that only 4 (20 percent) experienced recurrent tic pain. Dandy previously had commented on the frequent finding of vascular compression of the trigeminal nerve at the pons in patients with tic and in 1959 Gardner and Miklos reported the treatment of such vascular compression in one patient by moving the artery away from the nerve and maintaining the separation with a pledget of absorbable gelatin sponge. Jannetta has subsequently developed this useful technique and discusses its current status. Microvascular decompression of the trigeminal nerve through a retromastoid craniectomy is a nondestructive yet effective method of treating tic douloureux. It provides pain relief in a large percentage of patients without ordinarily causing trigeminal dysfunction, including anesthesia dolorosa. Like Dandy's suboccipital craniectomy, Jannetta's retromastoid approach would permit the discovery of a posterior fossa tumor that had not been detected preoperatively.

Comparison of Results of Surgical Treatment

In 1994, Taha and Tew analyzed their own results and those of major representative reports in the literature for the treatment of trigeminal neuralgia by radiofrequency thermocoagulation, glycerol rhizolysis, balloon compression, microvascular decompression, and partial trigeminal rhizotomy. Patients achieved a high incidence of initial pain relief after each of these procedures. However, microvascular decompression had the lowest rate of technical success in that 15 percent of patients underwent partial trigeminal rhizotomy instead, either because significant vascular compression was not found or adequate decompression could not be performed safely. Radiofrequency thermocoagulation and microvascular decompression had the highest rates of initial pain relief and the lowest rates of pain recurrence. Glycerol rhizolysis had the highest rate of pain recurrence. Balloon compression had the highest rate of trigeminal motor dysfunction. Balloon compression and microvascular decompression had the lowest rates of corneal anesthesia or keratitis. Microvascular decompression had the lowest rates of facial numbness and dysesthesia; all of the percutaneous procedures had similar rates of dysesthesia. Posterior fossa exploration had the highest rates of permanent cranial nerve deficit, intracranial hemorrhage or infarction, perioperative morbidity, and mortality.

A Therapeutic Approach to the Patient with Tic Douloureux

The acceptable approach to the treatment of tic douloureux is as follows: If the history, physical examination, and diagnostic studies show no evidence of a posterior fossa tumor, the patient is treated initially with Tegretol or neurontin. If these are unsuccessful, Dilantin is tried, alone or in combination with Tegretol or neurontin. Lioresal can be used as well, depending on the response to the first three medicines.

If medical therapy fails, the subsequent approach is based on the age of the patient and the location of the tic. In an elderly patient or a patient in poor health, with tic restricted to the forehead, supraorbital/supratrochlear nerve avulsion or balloon compression of the gasserian ganglion would be my next choice of treatment. In a similar patient with tic restricted to the cheek, alcohol injection or avulsion of the infraorbital nerve or one of the three percutaneous procedures (but favoring balloon compression to reduce the risk of corneal denervation). Finally, in an elderly or infirm patient with tic in the region of the eye, in the third division, or in multiple divisions, percutaneous trigeminal balloon compression, radiofrequency coagulation, or glycerol injection is recommended. These same recommendations apply to the patient of any age whose tic is related to multiple sclerosis.

If the patient is younger than 70 years, is in good health, and has not been helped or is no longer helped by medication, microvascular decompression is the treatment of choice, especially if the pain involves the area of the eye (less possibility of producing corneal anesthesia than with a percutaneous procedure) or the mandible (less likelihood of producing weakness of the muscles of mastication than with a percutaneous procedure). Intra­operative monitoring of brain stem auditory evoked potentials is used to minimize the risk of ipsilateral deafness. If definite vascular compression is not identified at the time of operation, the caudal half or two-thirds of the main sensory root of the trigeminal nerve should be divided adjacent to the pons. An alternative approach (which is actually the preferred choice of many authorities) in the healthy patient less than 70 years of age who has pain despite an adequate trial of medical therapy is percutaneous trigeminal balloon compression, radiofrequency thermocoagulation, or glycerol injection.

If pain recurs after any of these procedures, it should be approached therapeutically as a new event, beginning with a trial of medication. Alcohol injection, nerve avulsion, balloon compression, radiofrequency coagulation, gamma-knife and glycerol injection can be repeated, but recurrence after microvascular decompression is probably best approached by a percutaneous technique or by partial division of the sensory root at the pons rather than by a second microvascular decompression.

Whether it is to treat initial trigeminal neuralgia or recurrent trigeminal neuralgia, the neurosurgeon must bear in mind that there can be serious complications from the percutaneous procedures and from posterior fossa exploration. As reported by Sweet he and Poletti collected complication data from a number of neurosurgeons and found the incidence and variety of such un­toward events to be greater than expected based on reports published in the medical literature. As a general rule, the treating neurosurgeon will obtain the best results with the approach that he or she uses most frequently.