Introduction

Craniofacial pain is a common condition that affects approximately 10 to 25 % of the worldwide adult population, with a significant impact on quality of life. Twice as many women are affected with craniofacial pain as compared to men. Three of the most common causes of facial pain include trigeminal neuralgia (TN), trigeminal neuropathic pain (TNP), and persistent idiopathic facial pain (PIFP) [1]. These three conditions are complex pain syndromes that are often very difficult to manage with medications alone.

TN, or tic douloureux, is an extremely disabling illness, which was first described by Johannes Bausch in 1672 and later in 1773 by John Fothergill [2]. TN is capable of causing such intense suffering that it has sometimes been referred to as the suicide disease [3]. The worldwide peak incidence of TN is estimated to be approximately 4.5 cases per 100,000 population [4] and occurs in the sixth to eighth decades of life with a female predominance slightly higher than males in a ratio of 1.7:1.

According to the International Association for the study of pain, neuropathic pain is defined as “pain initiated or caused by a primary lesion or dysfunction in the nervous system.” Neuropathic pain can be further classified as peripheral or central (depending on the site of the disorder) and acute or chronic (lasting >3 months) [5]. While most neuropathic pain syndromes are characterized by allodynia, hyperalgesia, and background pain, TN typically lacks all of these features and can therefore be considered one of the purest forms of neuropathic pain [6].

Furthermore, TN is the most frequent and well-known hyperactive cranial nerve disorder, characterized by severe facial neuropathic pain in one or more divisions of the trigeminal nerve. The most common type of facial pain, also described as classic TN (Burchiel type 1), is associated with episodes of severe lancinating pain which occur either spontaneously or as a result of mild sensory or motor events [7]. Common triggers of TN pain include touching the face, brushing the teeth, wind, talking, and chewing. Aside from the typical lancinating pain of TN, the second most important finding in these patients is the absence or lack of background pain. This implies that between flairs of TN pain, these patients are essentially pain free [7].

TNP is defined as a constant burning, cramping, pricking, deep, aching, or electric shock-like facial pain along the distribution of the trigeminal nerve branches [6]. TNP is often associated with sensory dysfunctions such as paresthesia or dysesthesia, which can manifest as cold, pricking, tingling, or itching sensations [5]. While TNP can be accompanied by lancinating pain like TN, it is differentiated from TN because it is frequently associated with constant burning pain as well [7]. TNP can result from surgery, traumatic injury, and herpetic infection (shingles) of the areas innervated by the branches of the trigeminal nerve, including the sinuses, teeth, face, or skull [8].

The International Headache Society defines PIFP as “persistent facial pain that does not have the characteristics of the cranial neuralgias and cannot be attributed to other disorders” [9]. This condition is often described as severe persistent unilateral facial pain that is deep or poorly localized and usually burning or crushing in nature. There is usually a normal work-up without any demonstrable local cause for the persistent facial pain [10]. PIFP is usually not confined to any anatomic distribution and is often a diagnosis of exclusion [11].

Physiology of Facial Pain

Multiple mechanisms have been postulated in the pathophysiology of neuropathic facial pain, which accounts for a wide variety of clinical presentations [6]. The primary sensory innervation of the face is provided by the trigeminal system. Originating from the lateral pons, the trigeminal nerve splits into three divisions from the gasserian ganglion, namely the ophthalmic (V1), the maxillary (V2), and the mandibular (V3) divisions. Peripherally, the trigeminal nerve terminates as the supraorbital and supratrochlear nerves from V1, the infraorbital nerve from V2, and the mental nerve from V3. These areas consequently lend themselves to potential targets of neurostimulation [1].

The main sensory nervous supply of the posterior region of the head is provided by the greater, lesser, and third occipital nerves. Unlike the trigeminal nerve, the occipital nerves originate from the spinal nerves of the occipital plexus. The greater occipital nerve originates from the dorsal ramus of the C2 spinal root. The lesser occipital nerve is composed of branches of the C2 and C3 spinal roots [1]. Goadsby and Hoskin demonstrated that there is a connection between the trigeminal and cervical system [12]. Nociceptive afferent fibers from the trigeminal nerve and from the C2 to C3 region synapse with their second order neurons in the trigeminocervical complex (TCC), which projects from the trigeminal nucleus caudalis to the level of C3 [13]. Related to this confluence, the high cervical spinal levels are a common target for neurostimulation as a means of treatment for intractable headaches and chronic facial pain [14].

TN, TNP, and PIFP can be extremely difficult to treat with conventional pharmacologic therapy. Often times, patients are forced to suffer with this pain simply because no other treatment options exist. Microvascular decompression was proposed and developed by Peter J. Jannetta [15, 16]. It has become a standard surgical option for the treatment of TN. The procedure has been used with excellent success in patients where there is evidence of close proximity between elements of cranial nerve V and vascular structures, e.g., arteries and veins, which serve the brainstem and cerebellum and involves placement of a buffering pad between the portion of cranial nerve V and the offending artery.

The procedure is recommended for patients with inadequate medical control of pain, with greater than 5 years anticipated survival, and who are able to tolerate a small craniotomy and surgical anesthesia. Microvascular decompression generally produces pain relief in 90 % of patients with over 80 % remaining pain free for 1 year. Seventy-three to 75 % of patients remain pain free at 5 years after the procedure [15–20]. Complications including sensory loss, cerebral spinal fluid leaks, infarcts, and cranial nerve damage are infrequent, occurring less than 4–7 % of the time; the average mortality rate associated with the operation is 0.2–0.5 % [15–21]. A surgical variant of the Jannetta procedure was published in 2009, benefitting patients who during microvascular decompression were found intraoperatively not to possess a visible vascular compression. The procedure is called “nerve combing” and consists of splitting the nerve longitudinally into several fascicles [19, 20, 22].

Microvascular decompression is the only non-destructive surgical procedure for the treatment of TN. It also provides a low risk of facial sensory loss with subsequent dysesthesias or anesthesia dolorosa. Finally, microvascular decompression is the only operation which addresses what is believed to be the primary underlying pathology of vascular compression.

In addition to this surgical intervention, recent advances in neurostimulation may now offer novel less invasive approaches to treating a variety of craniofacial pain syndromes.

Peripheral Nerve Field Stimulation

Peripheral trigeminal nerve field stimulation was first introduced by Wall and Sweet in 1967 as a treatment modality for neuropathic pain [23]. Over the next half century, the use of peripheral nerve field stimulation (PNFS) in the treatment of intractable facial pain has become increasingly reported. Significant advances in electrode construction, generator manufacturing, and surgical technique now allow for the permanent implantation of trigeminal branch electrodes by using minimally invasive methods [24]. Most of the published literature on PNFS has reported significant improvement (>50 % on visual analog scales) in localized chronic pain intensity [1].

The surgical approach to PNFS is fundamentally based on the subcutaneous insertion of one or more electrodes into the painful area with subsequent electrical stimulation. Similar to spinal cord stimulation, a trial, usually lasting several days, is performed and the permanent system is then implanted in the case of a successful trial [25•].

Painful trigeminal neuropathy following craniofacial surgery or trauma, as well as post-herpetic trigeminal neuropathy, has been identified as indications well amenable to this neuromodulatory intervention, when more established therapies have failed [25•]. PNFS is a promising modality for the treatment of intractable facial pain; however, evidence still remains sparse.

In a recent retrospective study by Klein et al., a small cohort of ten patients with intractable facial pain underwent implantation of subcutaneous electrodes for trigeminal nerve field stimulation. Eight of these ten patients proceeded to implantation after a successful trial. Using the visual analog scale (VAS), average pain intensity was 9.3 preoperatively and 0.75 postoperatively. Furthermore, six patients reported complete absence of pain with stimulation and two had only slight constant pain without attacks [25•].

Although PNFS shows great promise for the treatment of facial pain, surgical implantation of these devices can often times be technically challenging, expensive, and cumbersome, prohibiting their widespread use. For the relief of pain in the head and face, several needs are easily identified. The implanting physician needs to have devices that are not linked to a large implantable pulse generator, which could eliminate tunneling, a major negative aspect of the procedure and a factor that leads to lead migration and decreased effectiveness of the implant [26••]. Possible solutions that are currently being developed include microwave technology [27] and external power sources [28].

Recently, a new eight-electrode wireless, microtechnology neuromodulation device has received FDA clearance for the relief of chronic back and leg pain. It is the world’s first wireless, fully programmable neuromodulation device [29•]. While there are currently no studies showing its effectiveness for chronic facial pain, considering its size (approximately 95 % smaller than conventional stimulation devices), these miniature electrodes may offer an easier and less invasive approach that may make them ideal for the treatment of facial pain.

PNFS for chronic refractory trigeminal neuralgia and trigeminal neuropathy of different etiologies may be an effective procedure when first-line therapies have failed [25•]. As it stands, stimulation systems are often “off label” when used for PNFS [24]. As such, prospective randomized controlled trials are still needed to further evaluate the use of PNFS in intractable facial pain syndromes [25•].

Sphenopalatine Ganglion Stimulation

Persistent idiopathic facial pain can be extremely difficult and significantly challenging to manage for the patient and the clinician. Pharmacological treatment of this painful condition is not always successful. The incidence of persistent idiopathic facial pain is 1/100,000; both sexes are affected equally, but more women than men seek medical care [10]. Widespread disagreement exists as to the pathogenesis of this chronic facial pain syndrome. There is some speculation that the pain is purely due to parasympathetic dysfunction that originates from the ganglion or from more complex central dysfunction [30].

The key structure in the expression of cranial autonomic symptoms is the sphenopalatine ganglion (SPG). It is theorized that the SPG may play a key role in the genesis of facial pain, and thus, electrical nerve stimulation of the ganglion would interfere with parasympathetic postganglionic outflow and therefore terminate neuropathic pain [30].

In a recent publication by Elahi et al., a case of an otherwise healthy 36-year-old female was described with a 9-year history of constant daily right-sided facial pain and significant episodic attacks. The patient reported the pain began insidiously and described it as deep, lancinating, sharp, and shooting. The pain was focused around the right maxillary, infraorbital region with some extension to the mandibular region. MRI showed no organic pathology. She had experienced the maximum tolerable doses of antiepileptic medications, antidepressants, and opioids with only minimal pain relief and significant side effects, including sedation, dizziness, and constipation [31•].

SPG block provided short-term pain reduction and so a trial of SPG nerve stimulation with percutaneous electrodes was attempted. The patient experienced a remarkably significant pain reduction during 5 days of the trial, and the decision was made for a permanent implantation. The patient reported great pain reduction at the 6-month follow-up with a reduction in chronic daily pain from 9 out of 10 to an average of 2 out of 10 on the NPR scale. She was able to completely wean off opioid medications [31•].

SPG stimulation may serve as a useful adjunct for those patients who fail conservative treatment for persistent idiopathic facial pain. The procedure, however, is not without risk. Potential pitfalls with this procedure may include injury to deep vessels, the facial nerve, and the parotid, as well as piercing bony structures. Even given the potential risks, this case report suggests a remarkable neuromodulatory role for SPG electrical stimulation in the treatment of medical refractory facial pain [31•].

High-Frequency Spinal Cord Stimulation

Spinal cord stimulation (SCS) is approved to treat chronic intractable pain of the trunk and limbs. SCS delivers electrical pulses via spinal epidural electrode arrays (leads) at vertebral levels associated with perceived pain [32]. Thoracic SCS is a well-established treatment modality for patients with failed back syndrome and complex regional pain syndromes [33].

Spinal cord stimulation in the high cervical (C1–2) region has fewer available studies; however, it has still been shown to be beneficial in treating both headache and facial pain syndromes [26••]. Initial trials of high cervical SCS were based on the prior success of nucleus caudalis dorsal root entry zone lesioning procedures for chronic pain, including trigeminal neuralgia. Stimulation over this confluence will potentially modulate pain in both the trigeminal and occipital distributions due to the geographic overlap of the fibers within the TCC [26••]. The procedure is usually performed in a similar manner to a standard thoracic spinal cord stimulator in two stages (trial followed by permanent implantation) [1].

In a distant case report by Barolat et al., they report the success of this therapy in relieving left-sided trigeminal neuralgia pain in a patient with advanced multiple sclerosis [34]. More recently, however, another study by Simpson et al., evaluating the efficacy of cervical spinal cord stimulation in 41 patients with intractable upper limb and face pain, concluded that patients with face pain did not respond to this therapy [35]. A recent retrospective study on the efficacy of cervicomedullary spinal cord stimulation in patients with TNP, post-herpetic neuralgia, trigeminal deafferentation pain (TDP), occipital neuralgia, and poststroke face pain concluded that a cervicomedullary junction stimulator is a favorable option for patients with TDP, TNP, and PHN, whereas those with occipital neuralgic pain rarely responded to this therapy [36]. Due to these conflicting results, more research is needed to elucidate whether high cervical spinal cord stimulation is a reasonable option for the treatment of facial pain. In this regard, recent advances in high-frequency spinal cord stimulation may offer yet another avenue as a treatment option for chronic facial pain.

In a recent randomized controlled trial comparing conventional spinal cord stimulation (40–60 Hz) to high-frequency HF10 stimulation (10 kHz), Kapural et al. found that HF10 therapy proved superior to traditional SCS in reducing back and leg pain over a 12-month follow-up period. Mean back pain VAS decreased 67 % with HF10 therapy compared with a 44 % decrease in VAS for traditional SCS. Mean leg pain VAS decreased 70 % over 12 months with HF10 therapy compared to a 49 % decrease with traditional SCS [32].

Currently, there are no randomized control studies evaluating the effectiveness of high-frequency SCS in the treatment of facial pain. Related to the superiority of high-frequency stimulation seen in treating chronic back and leg pain, further research should be pursued as to whether these same results can be duplicated for treating chronic facial pain at the high cervical levels.

Deep Brain Stimulation

The concept of relieving intractable pain with deep brain stimulation (DBS) appeared in the 1950s, a decade before the gate control theory [37]. DBS is defined as the use of electrical devices inserted intracranially and targeted at subcortical targets [26••].

DBS can be a last resort therapy for a variety of neuropathic and chronic pain syndromes that are refractory to medical therapy or other neuromodulatory techniques. DBS has gained significant popularity with improving efficacy, safety, and applications beyond movement disorders. DBS can be used for refractory pain syndromes such as neuropathic pain, deafferentation pain, brachial plexus avulsion pain, chronic low back pain, failed back surgery syndrome, and cluster headaches [38, 39].

Several patient series have shown benefits in stimulating various brain areas, including the sensory thalamus (ventral posterior lateral and medial), the periaqueductal and periventricular grey (PVG), and more recently, the anterior cingulate cortex [37]. However, this technique remains “off label” in the USA as it does not have Federal Drug Administration approval [37].

As reported in large case series of DBS for chronic pain, sensory thalamic stimulation has been used with varying effectiveness in several chronic pain syndromes with the VPM particularly targeted in facial pain [40]. A meta-analysis evaluating the efficacy of DBS for chronic pain reported that 56 % of the patients with neuropathic pain and none with nociceptive pain achieved long-term pain control with DBS of the sensory thalamic nucleus. Fifty-nine percent of patients with nociceptive pain and 23 % with neuropathic pain achieved long-term success with pain control following DBS of the PVG [39]. Overall, DBS of either the PVG/PAG or sensory thalamus has been shown to be more effective in alleviating nociceptive pain (63 %) than deafferentation/neuropathic pain (47 %) [39].

Although pain transmission pathways are well known, the exact mechanism by which DBS mitigates chronic pain still remains unclear. Nevertheless, several studies in animals and humans have demonstrated the PAG and thalamus as crucial structures of pain perception involved in chronic pain pathologies [41].

DBS for complex craniofacial pain syndromes may be indicated in patients who have failed other less invasive neuromodulatory techniques. However, even when DBS is truly efficacious, tolerance may manifest after several years. This may be overcome by slight changes in stimulation settings or by interrupting the stimulation periodically. Technological advances such as the so-called smart adaptive stimulation may allow patients to better control pain and reduce tolerance [37]. While DBS for pain has been shown to be effective in several patient series, prospective randomized controlled studies are still warranted to establish the efficacy of DBS therapy for complex craniofacial pain syndromes [1].

Conclusion

Complex craniofacial pain can be a challenging condition to manage both medically and surgically. With recent technological advances and success of neurostimulation therapy in small trials, there is a resurgence of interest in the role of neurostimulation therapy for the management of chronic facial pain [1]. Given the reversibility and minimally invasive nature of PNFS, this modality is being explored at a rapid pace and is preferred by both patients and surgeons [1].

Additionally, with the advent of new miniature wireless devices which may help eliminate some of the technical complexity associated with implantation, this may ultimately lead to the more widespread use of neurostimulation. To date, there is still a paucity of reliable evidence in the literature on the efficacy of sphenopalatine ganglion stimulation, spinal cord stimulation, and deep brain stimulation therapy for chronic craniofacial pain syndromes. In recent years, impetus has shifted towards the development of specialized, less invasive, miniaturized devices, which hope to maximize effectiveness of treatment, while minimizing risk [14].

The development of large, high-powered, adequately blinded trials is necessary to ensure the continued development of these promising methods. Many of the aforementioned trials and treatments under investigation for facial-related pain will become valuable additions to neuromodulation practices in the coming years [14].