Historical Development of Percutaneous Procedures for Trigeminal Neuralgia

Percutaneous procedures for treatment of trigeminal neuralgia were first introduced in 1853 by Patruban, who divided the maxillary nerve behind the orbit by passing a tenotome along the floor of the orbit and cutting the nerve as far back as possible.¹ In 1910, Harris applied a lateral approach to reach the foramen ovale and inject alcohol.² In 1914, Härtel modified the approach to an anterior one similar to that currently practiced,^3,4 and in 1930, Irger used an inferior approach (both approaches used alcohol injection).^4,5 Fortuitous serendipity led to the application of glycerol to treat trigeminal neuralgia. In 1981, Häkanson,⁶ while using glycerol as a medium to introduce tantalum dust into the trigeminal cistern as part of a stereotactic radiosurgical technique for the treatment of trigeminal neuralgia, noticed that patients became pain free, immediately. Later, he used glycerol alone and confirmed the pain relief effect.

It was Rethi who first introduced the use of electrolytic techniques for treatment of trigeminal neuralgia.⁷ In 1960, radiofrequency lesion was introduced by Sweet, and it was applied to the treatment of trigeminal neuralgia in 1974.^8,9 By applying Taarnhøj and Shelden’s^10–12 principle of internal neurolysis of the root of the trigeminal nerve in the dural canal over the margin of the petrous ridge, Mullan and Lichtor,^13,14 in 1983, introduced the technique of percutaneous balloon compression of the gasserian ganglion.

With the development of percutaneous techniques there was a need to ensure procedure safety. A first safety step was to ensure adequate needle placement to avoid extratrigeminal injury. This was first addressed in 1916 by Pollock and Potter¹⁵ in cadaveric studies, in which needle site during alcohol injection was confirmed using radiography. It was not until 1936 that Putnam and Hampton applied this procedure to patients, with good results.^10,16 A second safety step was to ensure physiologically adequate placement and adequate neurolysis without causing excessive damage. To this end, in 1955, Silverstone designed an insulated needle with an exposed tip used as a nerve stimulator.^10,17 Later, motor and sensory evoked potential monitoring was used in trigeminal destructive procedures by Karol and colleagues,¹⁸ and this was adopted by Sindou and coworkers.^19,20 Most recently, neuronavigation has been used to assist with percutaneous trigeminal neurolysis.^21,22

Trigeminal Anatomy

Percutaneous procedures in general aim to reach the trigeminal nerve ganglion or sensory root through the foramen ovale. An understanding of the relationship among the foramen, the third division of the nerve (mandibular), and surrounding structures is vital to avoid injury to the neighboring nervous system structures and vasculature.

The foramen ovale is situated in the anterior part of the sphenoid bone. It lies lateral to the foramen lacerum, which is occluded at its base only by cartilage. It is also near the posterior margin of the lateral pterygoid plate. Posterolaterally lies the foramen spinosum, which transmits the middle meningeal artery and meningeal branch of the mandibular nerve. Rarely, the foramen spinosum and ovale may be confluent. Posteroinferiorly lies the jugular foramen, and posterolaterally, the carotid canal.²³ The shape of the foramen ovale is typically oval, yet it can be almond shaped, round, or slit-like. The average length and width are 7.46 mm ± 1.41 mm and 3.21 mm ± 1.02 mm, respectively.²⁴ The foramen may be divided into two or three components in 4.5% of cases.²⁵

Several nerves, arteries, and veins pass through the foramen ovale: the mandibular nerve, lesser superficial petrosal nerve, accessory meningeal artery, and emissary veins (from the cavernous sinus to the pterygoid plexus). The optic ganglion is situated directly under the foramen but may also pass through the foramen ovale.²³

The gasserian ganglion originates from the cranial neural crest and the overlying thickened ectoderm. The three processes of the ganglion can be identified at 6 weeks’ gestation; the ophthalmic division develops first, followed by the maxillary and mandibular divisions. Meckel’s cave attains its final shape at 12 weeks’ gestation, with the arachnoid ending around the ganglia.²⁶

The gasserian ganglion is a sensory ganglion of the trigeminal nerve; it is crescent shaped with convexity directed forward. It is located at the apex of the petrous and may extend to the foramen lacerum, or the posterior lip or floor of the foramen ovale. The mean distance from the foramen ovale to the gasserian ganglion is 6 mm (range, 5.8 to 6.3 mm).²⁷ The anterior border of the ganglion is usually within 2 mm in front of, or behind, the posterior lip of the foramen ovale.²⁸ The ganglia’s dural covering forms Meckel’s cave and, medially, the cavernous sinus.²⁸

Meckel’s cave is confined between the outer and inner layers of the dura but also contains a separate sleeve of meningeal dura formed by extension of the posterior fossa dura into the middle fossa. Surfaces are concave inferiorly and medially and more flat superiorly. The sensory branches of the trigeminal nerve and partially the gasserian ganglia and its three divisions occupy Meckel’s cave. The subarachnoid space within Meckel’s cave extends an average of 4.9 mm medially and 1.7 mm laterally beyond the posterior edge of the gasserian ganglia. This space may extend over the divisions of the trigeminal nerve. Mean length of Meckel’s cave ranges from 6 to 16 mm.²⁹ The gasserian ganglion measures 4 to 5 mm wide and 15 to 25 mm long. The length of the sensory root varies from 5 to 15 mm; the length of the mandibular nerve from the anterosuperior margin of the foramen ovale to the gasserian ganglion is 0 to 10 mm.^28–30

The ganglion has a posterior sensory root, which joins the brainstem about halfway between the lower and upper borders of the pons. It starts with an oblique run upward from the lateral part of the pons toward the petrous apex, exits the posterior fossa, enters the middle cranial fossa by passing forward beneath the tentorial attachment, and enters Meckel’s cave to join the ganglia through a deep hilum on its posterior aspect. The motor branch of the trigeminal nerve runs in front of and medial to the sensory root and passes beneath the ganglion, leaving the skull through the foramen ovale and, immediately below this foramen, joining the mandibular nerve.³¹

The sensory rootlets, between the gasserian ganglion and the root in the cerebellopontine angle, form a plexus at its entry into the ganglion. This retrogasserian triangular area has been named the triangular plexus and has been identified as the best place to create a lesion for the treatment of trigeminal neuralgia.³²

The trigeminal ganglion is formed by union of the three divisions of the trigeminal system: the mandibular nerve, maxillary nerve, and ophthalmic nerve, also known as V3, V2, and V1, respectively. The ophthalmic nerve is the smallest of the three trigeminal divisions. It inclines upward as it passes forward near the medial surface of the dura, forming the lower part of the lateral wall of the cavernous sinus and reaching the superior orbital fissure. The maxillary nerve takes a more direct course and enters the foramen rotundum. The mandibular nerve occupies most of the gasserian ganglion and takes a caudolateral course from the ganglion and enters the foramen ovale.^31,33

Diagnosis

The first known references to what is believed to be trigeminal neuralgia were by Aretaeus of Cappadocia³⁴ and later by Avicenna, who described trigeminal neuralgia in his text Canon Medicinx.^35,36

The International Association for the Study of Pain (IASP) defined the essential features of idiopathic trigeminal neuralgia as “sudden, transient, intense bouts of superficially located pain, strictly confined to the distribution of one or more divisions of the trigeminal nerve usually precipitated by light mechanical activation of a trigger point or area.”³⁷ Penman, in a detailed description of trigeminal neuralgia, stated that the only constant feature is the symptom of pain that is unilateral, trigeminal, severe, paroxysmal, and precipitated.³⁸

A diagnosis of trigeminal neuralgia is dependent on the patient’s history and analysis of the patient’s complaint. Pain is usually described as spreading outward from the trigger point to cover an area that roughly approximates the territory of distribution of one (or more) of the trigeminal divisions.³⁹ Usually, there is no clinically evident sensory deficit, yet quantitative and qualitative tests have demonstrated deficits for tactile and warm sensations in the affected area, with no deficit for heat pain or for pinprick.^39–42 There is also an absolute, followed by a relative, refractory period after an attack, during which further paroxysms cannot be elicited by trigger-point stimulation.⁴³

A differential diagnosis of trigeminal neuralgia includes a large array of facial pains. Trigeminal neuralgia itself may be due to multiple causes and pathologic processes, which influence the mode of treatment and outcome. To assist in disease management, a trigeminal neuralgia classification system has been proposed by Burchiel.^44,45

Diagnostic Tools

Trigeminal evoked potentials and electrophysiologic studies are not widely used but are complementary diagnostic tools.^46,47 Magnetic resonance imaging (MRI) and magnetic resonance angiography (MRA) processed as three-dimensional images have been used to verify vascular compression.^48–50 MRI may not be necessary in cases managed purely by a percutaneous procedure.

Surgical Treatment

There are no specific surgery guidelines for patient selection or guidelines for which surgical method to select. A recent report details the diagnostic evaluation and treatment of trigeminal neuralgia as an evidence-based review and makes some recommendations.⁵¹

Surgery is generally offered to patients who fail medical treatment, have only partial relief of pain after 1 year, or according to Garvan and Siegfried,⁵² still require medication after consumption of more than 3000 tablets of a single drug. Microvascular decompression (MVD) remains the surgical option of choice for trigeminal neuralgia, with percutaneous procedures generally reserved for patients who experience recurrent pain after MVD, are a high surgical risk owing to medical comorbidity, and are older than 65 to 70 years of age, although this latter indication has been questioned.^53,54 Percutaneous procedures may also be offered to patients with multiple sclerosis (MS).^55–57 In a case of suspected symptomatic trigeminal neuralgia⁵¹ due to MS, further evaluation may be needed for surgical planning and prognostication.

Percutaneous Preoperative Care

All percutaneous procedures for the management of trigeminal neuralgia involve producing some injury to trigeminal afferents. This is achieved by using heat in the case of radiofrequency rhizotomy, chemical neurolysis in the case of glycerol rhizotomy, mechanical neurolysis in the case of balloon compression rhizotomy, and radiation-induced neural injury in the case of stereotactic radiosurgery (SRS). That so many percutaneous procedures are available points to the fact that none have proved ideal. That is, no one percutaneous procedure is applicable in all cases with a uniformly high rate of long-term success and with minimal possibility of complication or recurrence. Radiofrequency rhizotomy is perhaps the most commonly used procedure, and long-term case results are available.^57–59

Percutaneous Surgical Procedures

Radiofrequency Rhizotomy

Radiofrequency (RF) rhizotomy was popularized by Sweet and Wepsic, as reported in 1974.⁹ The concept behind the technique is to use radiofrequency stimulation (alternating electric field with an oscillating frequency of 500,000 Hz) to cause a thermal lesion in the retrogasserian root, or ganglion.⁶⁸ In an animal study, radiofrequency heating was reported to cause selective injury to small myelinated and unmyelinated fibers⁶⁹; however, a subsequent neuropathologic clinical model study failed to substantiate this.⁷⁰ It is now generally believed that radiofrequency heating causes a nonselective destruction of axons, regardless of fiber size.⁷¹

In the operating suite, the patient is placed supine on the operating table. After intravenous access is secured, anticholinergics (usually 0.4 mg of atropine) may be given to minimize oral secretions and blunt the vasovagal response during penetration of the needle through the foramen ovale. Some surgeons prefer to watch for the vagal response as a guide while engaging the foramen ovale and may not use preoperative anticholinergics. If anticholinergics are not used, an external pacemaker strapped to the chest wall, set to deliver 45 beats per minute, should bradycardia occur during electrode insertion or during stimulation, may be used.⁷¹

Anesthesia is induced using a short-acting agent, such as propofol, supplemented with a parenteral short-acting opiate analgesic (e.g., alfentanil). The patient is then placed with the head extended to about 30 degrees to obtain a clear submental-vertex view of the foramina ovale and spinosum, located on the greater wing of the sphenoid bone, of the affected side. Biplanar fluoroscopy can be used if available. Usually a portable C-arm digital fluoroscope is sufficient. Magnification (2×) of the fluoroscopic image may be required to visualize the foramen clearly.

The skin of the cheek and perioral region on the ipsilateral side is painted with antimicrobial solution after the required depth of anesthesia has been obtained (eyelash reflex is lost). A nasal cannula or “trumpet” may assist in the maintenance of the airway. The surgeon stands on the same side of the patient’s pain, and cannulation of the foramen proceeds under fluoroscopic guidance using the Härtel technique with the RF electrode trochar.^3,72 The needle enters the cheek through a point 2.5 cm lateral to the corner of the mouth and 1 cm inferior to the occlusal plane and advances through the cheek in the submucosal layer, between the pterygoid bone and the angle of the mandible, guided by the surgeon’s index finger of the other hand within the patient’s mouth (Fig. 161-1). The needle is directed along a line representing the intersection of a vertical plane passing through the lateral aspect of the ipsilateral pupil and a horizontal plane passing through a point 3 cm anterior to the external auditory meatus along the inferior border of the zygoma. The needle engages the foramen ovale about 6 to 8 cm from the skin surface. This can usually be evidenced by a contraction of the masseter muscle manifested as jaw jerk (see Fig. 161-1).

FIGURE 161-1 The radiofrequency needle is inserted at a point 2.5 cm from the corner of the mouth, 1 cm below the occlusal plane. The trajectory is the intersection of two planes: one directed to the mid-pupil, and the other to a point 3 cm anterior to the tragus. During insertion, the index finger is inserted lateral to the teeth to prevent intraoral penetration and to ensure that the needle is not directed lateral to the maxilla.

Lateral fluoroscopy is used thereafter, and the needle is advanced to a point just superior to the intersection of petrous bone and the clivus on a true lateral view. The patient is now awakened from anesthesia, and the stylet of the trochar is withdrawn and replaced with the RF electrode. A curved-tip electrode (Tew) may be necessary to target V1 and V2, but typically a straight electrode suffices to produce a lesion in V3 (Fig. 161-2).

FIGURE 161-2 On lateral fluoroscopy, the needle is seen entering the foramen ovale, directed to a region that is just anterior to the intersection of the petrous bone and clivus. For V3, the needle is directed at the 10-o’clock position, and for V2/V1, the needle is directed more superiorly toward the 11- or 12-o’clock position. The latter will require a somewhat lower and more lateral entry point on the cheek.

Once the patient is adequately awake, test stimulation of the target is achieved using a square wave stimulus of 0.05 to 0.15 volts, at 50 Hz, using a 1-millisecond pulse duration, to evoke paresthesias in the stimulated division. A somewhat higher threshold of stimulation may be required in patients who have had a previous lesioning procedure. When the electrode location has been confirmed, the patient is again anesthetized. Lesion creation is achieved by heating the electrode up to 75° to 80° C for 90 seconds. For lesions in the V1 segment, a duration of 60 seconds may be used. The initial lesion may be “bracketed” by making a lesion on either side by moving the electrode 2 mm proximally and distally from the initial position. After neurolysis is complete, the patient is again awakened from anesthesia to test the procedure efficacy. The end point of this procedure is slight hypoesthesia, such that the patient is able to feel touch in the affected area but cannot differentiate between the sharp and blunt ends of a safety pin.

When the desired end point has been achieved, the needle with the electrode is withdrawn, and the entry point is checked for bleeding. Rarely, pressure application may be required in the presence of bleeding to preempt the formation of a hematoma. The patient is then transferred to the recovery room and is kept under observation for a few hours. Typically, the patient can return home the same day.

In 98% to 99% of the patients, pain relief is obtained immediately after surgery.^73,74 A recurrence rate of 20% at 9 years has been described, with an incidence of 9% for mild dysesthesia, 2% for major dysesthesia, and 0.2% for anesthesia dolorosa.⁷³ Because hypoesthesia is the end point of the procedure, numbness is present in 100% of patients after surgery.⁷³ The irreversible loss of long latency trigeminal root evoked potentials has been documented as a means of objective assessment of the effects of the rhizotomy,⁷⁵ and a recently described multiarray electrode method for mapping the trigeminal nerve may help in producing more selective lesions.⁷⁶