Atypical neuropathic facial pain is a syndrome of intractable and unremitting facial pain that is secondary to nociceptive signaling in the trigeminal system. These syndromes are often recalcitrant to pharmacotherapy and other common interventions, including microvascular decompression and percutaneous procedures. Herein, the authors present two other viable approaches (nucleus caudalis dorsal root entry zone lesioning and motor cortex stimulation), their indications, and finally a possible treatment algorithm to consider when assessing patients with atypical facial pain.
Key points
- •
Atypical facial pain is often recalcitrant to pharmacotherapy; appropriate patient selection can make surgery a viable and appropriate option for treatment.
- •
Nucleus caudalis dorsal root entry zone lesioning is an effective and safe means of treating atypical facial pain.
- •
Motor cortex stimulation is an alternative intervention to consider for patients who do not respond to nucleus caudalis dorsal root entry zone lesioning.
Introduction
Facial pain syndromes encompass a body of diagnoses that can fit a very defined and classic presentation to one that is murkier and not localizable by physical examination or radiographic imaging. Beyond typical trigeminal neuralgia, atypical subtypes are often difficult to treat and refractory to common interventions. Herein, the atypical subtype is used as a catchall phrase to define trigeminal pain, not defined by trigeminal neuralgia type I or II by the Burchiel classification ( Table 1 ). The current spectrum of treatment of these syndromes includes pharmaceutical therapy, percutaneous procedures (chemical, thermal, irradiative, mechanical destruction), as well as open surgery, which includes microvascular decompression (MVD) and less often peripheral nerve stimulation, motor cortex stimulation, neurectomy, tractotomy, and nucleus caudalis dorsal root entry zone (NC DREZ) lesioning, though there is currently no mutual understanding of terminology, diagnosis, or treatment of facial pain ( Table 2 ). Several treatment algorithms have been developed for the treatment of facial pain with these various modalities. Alksne and colleagues used preoperative risk assessment and the presence of a vascular loop on MRI imaging for the treatment of trigeminal neuralgia with MVD versus Gamma Knife radiosurgery. In the single-institution experience by Munawar and colleagues atypical subtypes as well as less-often-used treatment modalities are incorporated in a more comprehensive treatment algorithm. Specifically, treatment with tractotomy is considered only in cases of cancer facial pain. Neurectomy and motor cortex stimulation are used in cases of trigeminal neuropathic pain and deafferentation-type injuries. Of note, NC DREZ lesioning is considered less often and thought to be a more morbid and invasive procedure. Since 1982, NC DREZ lesioning at Duke has been refined to become safer and less invasive. Furthermore, patient outcomes in the Nashold experience were reassuring with 96% pain relief postoperatively and 67% even still at 1 year. Patients were treated for postherpetic neuralgia, deafferentation pain (anesthesia dolorosa, post-tic dysesthesia, stroke, multiple sclerosis, gasserian tumor, Gamma Knife radiation injury), facial trauma/surgery, atypical facial pain, and migraine/cluster headache with success. More recently, Chivukula and colleagues reported 75% improvement in quality of life and greater than 50% pain reduction at the 8-year follow-up (16 patients). Therefore, the authors emphasize NC DREZ for the treatment of these atypical subtypes, a procedure the authors think is underused and, in their cohort, proven to be very effective. With this experience in mind, the authors consider a new treatment algorithm for the treatment of atypical facial pain.
| Pain Category | History | |
|---|---|---|
| Typical | Trigeminal neuralgia (type 1) | <50% episodic, typical trigeminal pain |
| Trigeminal neuralgia (type 2) | >50% constant, typical trigeminal pain | |
| Atypical | Trigeminal neuropathic pain | Unintentional injury secondary to dental/sinus injury, craniofacial trauma |
| Trigeminal deafferentation pain (anesthesia dolorosa) | Intentional deafferentation secondary to destructive procedure | |
| Symptomatic trigeminal neuralgia | Multiple sclerosis, mass lesion (posterior fossa lesion) | |
| Postherpetic neuralgia | Herpes zoster (trigeminal distribution) | |
| Atypical facial pain | Somatoform pain disorder |
| Procedure | Details | Notes | |
|---|---|---|---|
| First order neurons | Microvascular decompression | Retrosigmoid craniectomy and vascular decompression of offending vessel from trigeminal nerve root | Nondestructive, complications include cerebrospinal fluid leak and vascular injury though routine procedure, widely accepted |
| Destructive lesioning | Percutaneous needle insertion; chemical (glycerol), thermal (radiofrequency), or mechanical (balloon) destruction of trigeminal nerve | Destructive, does not require general anesthesia, can be selective for pain distribution, may result in numbness or dysesthesia, immediate pain relief | |
| Stereotactic radiosurgery | Stereotactic radiation of trigeminal nerve root | Destructive, noninvasive, may take several months for symptom relief | |
| Peripheral nerve stimulation | Peripheral trigeminal nerve electric stimulation | Nondestructive, trial period before permanent placement, adjustable settings for optimization | |
| Second order neurons | NC DREZ lesioning | Radiofrequency ablation of nucleus caudalis from a posterior midline approach | Destructive, invasive, risk of transient ipsilateral limb ataxia |
| Third order neurons | Motor cortex stimulation | Epidural stimulation of motor cortex through burr hole | Nondestructive, invasive requiring craniotomy, adjustable settings for optimization |
Diagnosis and Anatomy
Standard inclusion and exclusion criteria do not exist for either NC DREZ or motor cortex stimulation. A comprehensive physical examination, history, and evaluation both by a pain neurologist and psychologist (to exclude somatoform disorder) are necessary for the workup of chronic recalcitrant facial pain. Typical trigeminal neuralgia is often described as episodic, sharp, and shocklike in the trigeminal distribution occurring in bursts either less than 50% (type 1) or greater than 50% (type 2) of the time, triggered by non-noxious stimuli, such as speaking, eating, or brushing teeth. Atypical facial pain, on the other hand, is different in quality, more often described as constant, burning, throbbing, and aching facial pain. Furthermore, this type of pain is characteristically unresponsive to anticonvulsant drugs. At the time of surgical consideration, other causes, such as typical trigeminal neuralgia and headache, must be ruled out. Finally, psychological evaluation and testing is imperative.
Anatomically, trigeminal afferents ( primary neurons ) carrying pain and temperature fibers enter the pons and bifurcate, sending a caudal branch into the medulla, known as the descending trigeminal tract. At the level of the cervicomedullary junction, the spinal trigeminal nucleus overlies the descending tract. This tract also carries with it sensory fibers from the 7th, 9th, and 10th cranial nerves. The trigeminal afferents are topographically separated into 3 subdivisions (rostral to caudal): nucleus oralis, nucleus interpolaris, and nucleus caudalis located at the cervical medullary junction extending down to the level of C2. Secondary neurons then go on to synapse in the ventroposteromedial nucleus before tertiary neurons terminate at the level of the somatosensory cortex. Deafferentation injury is thought to lead to spontaneous signaling of secondary neurons leading to central facial pain. In this article, the authors consider therapies that target nociceptive signaling at the level of the trigeminal nucleus caudalis (NC DREZ), extending to the somatosensory cortex (motor cortex stimulation). Furthermore, the authors differentiate therapies as those targeting primary neurons (MVD, destructive lesioning, stereotactic radiosurgery), secondary neurons (NC DREZ), and tertiary neurons (motor cortex stimulation) of the trigeminal pain pathway.
Surgical Options
Nucleus caudalis dorsal root entry zone lesioning
Ablation of the trigeminal pain pathway by way of tractotomy-nucleotomy or by open surgical NC DREZ lesioning has been shown to be an effective means of treating atypical facial pain. First operated on by Sjöquist, several modifications have been made since to trigeminal tract surgery. Typically, tractotomy and nucleotomy are considered before NC DREZ. Although the data comparing these procedures are limited, reported benefits for NC DREZ seem to be superior to tractotomy and nucleotomy for facial pain. Specifically, NC DREZ is effective for postherpetic neuralgia and craniofacial pain conditions and has also been used for chronic cluster headache, vagal or glossopharyngeal neuralgias, and intractable pain syndromes secondary to cancer/craniofacial surgery/trauma. This is because pain fibers from not only the area of the trigeminal nerve but also the facial, glossopharyngeal, and vagus cranial nerves synapse in second order neurons in the nucleus caudalis. The procedure uses the fact that pain- and temperature-carrying fibers separate from motor and touch sensation at the pons, therefore, allowing selective ablation. These first order neurons descend as part of the descending trigeminal tract and synapse to secondary neurons of the nucleus caudalis where it is thought spontaneous pain generation occurs in the case of deafferentation injury.
NC DREZ can use computed tomography (CT)–guided stereotactic navigation as first described by Kanpolat and colleagues for tractotomy. A mini vertical midline incision is centered over the foramen magnum and dissected down to the C1 arch to ultimately expose dura. With the cord exposed ( Fig. 1 ), relevant anatomy can be visualized, including the vertebral artery, posterior inferior cerebellar artery, as well as the accessory nerve rootlets, C1/C2 roots, and pial vessels. Endoscopic assistance is also used for further visualization of anatomy beyond the extent of our exposure. Somatosensory-evoked potentials are then used to ensure trigeminal afferents are targeted as well as motor evoked potentials to ensure absence of the corticospinal tract. Given the proximity of the accessory nerve rootlets, trapezius stimulation can be seen. Once anatomic and electrographic anatomy is ensured, lesions are made to the nucleus caudalis extending from the obex to the level of the C2 nerve root. Between 1982 and 1988, in its first iteration, NC DREZ lesioning resulted in transient ipsilateral limb ataxia as a result of damage to the neighboring spinocerebellar tract. These side effects of the procedure eventually resolved but led to the second phase of the procedure with new electrode design incorporating insulation near the tip, which spared the spinocerebellar tract. Finally third-generation electrodes were developed to account for nucleus caudalis anatomy. Specifically, the cephalad nucleus caudalis is larger (∼2 mm) and, below the level of C1, diminishes in size and becomes elongated and elliptical. As a result, 2 electrode tips were developed to lesion a deeper/wider nucleus caudalis at the level of the obex and a shallower/narrower nucleus caudalis between C1 and C2. Finally, in its current form, the authors have used CT/MRI guidance, trigeminal neuromonitoring, and a minimally invasive endoscopic-assisted lesioning. A post-operative MRI is shown in Fig. 2 , demonstrating the area of lesioning.
The authors’ experience at Duke over the last 3 decades with NC DREZ has been positive. With greater than 70% pain relief, transient complications, and very high patient satisfaction, the authors treat atypical facial pain syndromes with DREZ lesioning as opposed to peripheral nerve stimulation, neurectomy, and motor cortex stimulation. Given its minimal invasiveness, safety, and reproducibility, the authors also advocate for its consideration as a therapeutic option for these challenging conditions refractory to other medical and interventional therapies.
Motor cortex stimulation
Penfield first demonstrated the idea that cortical stimulation can produce analgesia in the 1930s. Not until 1991 did Tsubokawa and colleagues formally describe motor cortex stimulation as a treatment of facial pain. Mechanistically, it is thought that motor cortex stimulation restores highly organized reciprocal pathways between primary motor and somatosensory cortex (layer I), and it is known that the somatosensory cortex can attenuate nociceptive signals. Deafferentation injuries can disrupt this attenuation leading to aberrant connections; therefore, nonpainful stimuli elicit severe pain. Other mechanisms suggest activation of functional networks involving the anterior cingulate cortex, orbitofrontal cortex, medial thalamus, and periaqueductal gray (PAG). Finally, some evidence suggests that motor cortex stimulation increases secretion of endogenous opioids particularly in the anterior middle cingulate cortex as well as PAG. Although motor cortex stimulation seems to alleviate deafferentation injury–related pain, no standard inclusion or exclusion criteria exist for treatment.
The procedure entails introduction of an electrode following burr hole placement. This procedure is done epidurally, though subdural lead placement has also been described. Image-guided neuronavigation is also used to ensure that the lead is placed over the premotor and precentral gyrus. Since its initial inception, a limited number of studies have described the efficacy of this procedure in a variety of pain syndromes, including poststroke pain, phantom-limb pain, postherpetic neuralgia, and other neuropathic facial pain syndromes. In fact, of these studies, 75% to 100% of patients reported good or excellent pain relief. Specifically, in a study by Ebel and colleagues, 7 patients with trigeminal neuropathic pain (dysesthesia, anesthesia dolorosa, and postherpetic neuralgia) underwent motor cortex stimulation. In all but one patient, a successful test stimulation period led to permanent lead placement. Of those patients, 50% maintained excellent pain control at the 5-month to 2-year follow-ups. A subsequent and more comprehensive report of 10 patients by Brown and Pilitsis again showed that 7 of 8 patients who underwent permanent lead placement had immediate pain relief postoperatively of 50% or greater. Seventy-five percent of these patients continued to have sustained pain relief at the 3- to 24-month follow-up. A more recent review of the literature identified 118 patients with chronic neuropathic facial pain treated with motor cortex stimulation. Of these cases, 84% continued to have good pain relief by the end of their study period. Of the whole reviewed group, the most common complication was seizure followed by wound infection. Of note, 2 patients treated by Henderson and colleagues were able to recapture the benefit of motor cortex stimulation with reprogramming of the stimulator parameters. In fact, a review of these stimulation parameters showed a wide variety of rate (hertz), pulse width (microseconds), and amplitude (milliampere), without a clear consensus. Therefore, although motor cortex stimulation has been shown to be effective in managing patients with common treatment-refractory neuropathic facial pain, there may be room for even further improvement by optimizing stimulation parameters.
Other surgical interventions
Tractotomy and nucleotomy, first described in 1938, targets facial pain at the tractus at the cervicomedullary junction. In a series of 17 patients, Kanpolat and colleagues performed this procedure with excellent pain relief in 44% and moderate pain relief in 56% of patients. One patient remained with pain and underwent NC DREZ, with subsequent resolution of pain relief up to the 2-year follow-up. Given the minimally invasive nature of the procedure, it was, therefore, advocated that patients undergo tractotomy-nucleotomy before NC DREZ is considered.
Since its inception in the middle of the twentieth century, deep brain stimulation of the thalamus and caudate has also been used in the treatment of chronic pain. However, results of 2 multicenter trails were unfortunately inconclusive. Several stimulation targets have been used, including PAG, zona incerta, and the ventral posterolateral nucleus of the thalamus. In a recent double-blinded study by Rasche and colleagues of 56 patients with neuropathic pain, 6 patients had trigeminal neuropathic pain. Of these patients, half, at 30 months of mean follow-up, had an unsatisfactory response to the treatment. Deep brain stimulation, however, remains a poorly understood intervention for neuropathic pain; much remains to be learned about its role in the facial pain treatment algorithm.
Related posts:
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
