INTRODUCTION — Neck pain and cervical muscle tenderness are common and prominent symptoms of primary headache disorders [1]. Conversely, it is plausible that head pain can be referred from bony structures or soft tissues of the neck, a condition called cervicogenic headache [2]. Nevertheless, there has been debate regarding the existence of cervicogenic headache as a distinct clinical disorder, as well as its underlying pathophysiology and source of pain [3-7].
This topic will review the definition, etiology, diagnosis, and treatment of cervicogenic headache.
ETIOLOGY AND ANATOMY — There is evidence that headache can be caused by referred pain from the upper cervical joints [6].
The anatomic locus for cervicogenic headache is the trigeminocervical nucleus in the upper cervical spinal cord, where sensory nerve fibers in the descending tract of the trigeminal nerve (trigeminal nucleus caudalis) are believed to interact with sensory fibers from the upper cervical roots (figure 1) [8]. This functional intersection of upper cervical and trigeminal sensory pathways is thought to allow the bidirectional transmission of pain signals between the neck and the trigeminal sensory receptive fields of the face and head.
The first three cervical spinal nerves and their rami are the primary peripheral nerve structures that can refer pain to the head (figure 2):
●The C1 spinal nerve (suboccipital nerve) innervates the atlanto-occipital joint. Pathology or injury affecting this joint is a potential source for pain that is referred to the occipital region of the head.
●The C2 spinal nerve and its dorsal root ganglion have a close proximity to the lateral capsule of the atlanto-axial (C1-2) zygapophyseal joint and innervate the atlanto-axial and C2-3 zygapophyseal joints. Trauma to, or pathologic changes around, these joints can be a source of referred head pain.
●The third occipital nerve (dorsal ramus C3) has a close anatomic proximity to, and innervates, the C2-3 zygapophyseal joint. Pain from the C2-3 zygapophyseal joint is referred to the occipital, frontotemporal, and periorbital regions of the head ("third occipital headache").
Involvement of the C2-3 zygapophyseal joint is the most frequent source of cervicogenic headache, accounting for up to 70 percent of cases [6,9-13]. The evidence implicating this joint as the source of cervicogenic headache is derived from studies of patients with a history of neck injury or trauma [6]. As an example, one series of 100 patients who had chronic neck pain following whiplash found that the prevalence of third occipital headache was 27 percent [9].
Involvement of the atlanto-axial joint is probably the second most common source of cervicogenic headache, but the true frequency is unknown [6,11]. Patients with rheumatoid arthritis are prone to subluxation and arthropathy of the atlanto-axial joint. Uncommon sources of cervicogenic headache include the C3-4 zygapophyseal joint, upper cervical intervertebral discs, and lower cervical zygapophyseal joints [10,14,15].
EPIDEMIOLOGY — Available data suggest that the prevalence of cervicogenic headache in the general population is 0.4 to 4 percent [16-19]. However, there have been no rigorous, large scale epidemiologic studies.
DEFINITION AND CLINICAL FEATURES — There is no global consensus regarding the definition of cervicogenic headache, but two conflicting viewpoints are widely recognized [20]:
●One approach employs a definition that relies on clinical features, as proposed in 1983 with introduction of the term "cervicogenic headache" [2]. In this sense, cervicogenic headache is characterized by unilateral head pain of fluctuating intensity that is increased by movement of the head and radiates from occipital to frontal regions [2,3,21,22]. The headache is precipitated by neck movement or sustained awkward head positioning. The headache does not shift from side to side. The pain is typically nonthrobbing, nonlancinating, of moderate to severe intensity, and of variable duration. Patients with cervicogenic headache may have restricted neck range of motion, and may have ipsilateral neck, shoulder, or arm pain.
One problem with this definition is that the proposed clinical features of cervicogenic headache overlap with those commonly associated with primary headache disorders such as tension-type headache, migraine without aura, and hemicrania continua. Many primary headache disorders may have associated neck pain [23]. As a result, distinguishing cervicogenic headache from other headache types can be difficult. Additionally, pathologies that may result in cervicogenic headache (eg, spondylosis, inflammation, and/or trauma) may affect both sides. As such, while symptoms should be asymmetric, they may not be unilateral. Given these concerns, the validity of the proposed diagnostic clinical criteria has been challenged. (See 'Diagnostic criteria' below.)
●Another approach ignores clinical features of cervicogenic headache and relies instead on establishing the diagnosis by demonstrating a cervical source of head pain and confirming the diagnosis by anesthetic blocks that pinpoint sources of pain in the upper cervical joints [20]. (See 'Diagnostic anesthetic blockade' below.)
DIAGNOSIS — Anesthetic blockade of the lateral atlanto-axial joint, the C2-3 zygapophyseal joint, the C3-4 zygapophyseal joint, and/or less commonly the mid-lower cervical zygapophyseal joints can be used to confirm the diagnosis of cervicogenic headache. (See 'Diagnostic anesthetic blockade' below.)
Some investigators believe that the diagnosis of cervicogenic headache can be made in certain cases without doing an anesthetic procedure, if the diagnosis is based on a careful history, physical examination, and a complete neurologic assessment. Limited evidence suggests that the presence of a combination of physical measures such as palpably painful upper cervical joints, restricted range of neck extension, and muscle impairment characterized by reduced electromyographic activity in the deep neck flexors can distinguish cervicogenic headache from migraine and tension-type headache [24]. However, the diagnostic utility of this method has not been validated, and its use requires specialized skills and equipment. Additionally, patients with neck pain from zygapophyseal arthropathy (with or without associated cervicogenic headache) often develop a secondary myofascial neck pain as a consequence of restricted neck range of motion. As such, pain with palpation in the region of a zygapophyseal joint may not discriminate a zygapophyseal from a myofascial pain generator.
Others argue that cervicogenic headache cannot be diagnosed on clinical grounds alone since the definitive criterion is complete relief of pain after controlled anesthetic blocks of cervical structures or their nerve supply [25]. In support of the latter viewpoint, a study that evaluated 71 patients with chronic neck pain and headache found no distinguishing history or examination features that confirmed a definitive diagnosis of third occipital headache before nerve blocks [9].
Since accurate diagnosis of cervicogenic headache is a prerequisite to appropriate treatment, we suggest that the diagnosis should be confirmed by controlled anesthetic blocks of cervical structures or their nerve supply. These blocks should be performed by an experienced specialist with imaging guidance, as discussed below. (See 'Diagnostic anesthetic blockade' below.)
Other headache etiologies should be pursued if the response to diagnostic blockade is incomplete.
Imaging findings in the upper cervical spine are common in patients without headache. While cervical spine imaging with plain films, magnetic resonance imaging (MRI), computed tomography (CT), and/or CT myelography may identify candidate structures that could cause cervicogenic headache (eg, asymmetric facet arthropathy or rheumatoid changes around the atlanto-axial [C1-2] joint), they cannot confirm the diagnosis of cervicogenic headache [26,27]. Imaging is primarily used to identify candidate pathology and to search for secondary causes of pain that may require surgery or other, more aggressive forms of treatment [28].
Diagnostic criteria — Diagnostic criteria for cervicogenic headache have been proposed by two groups: the Cervicogenic Headache International Study Group (CHISG) [22] and the International Classification of Headache Disorders, 3rd edition (ICHD-3) [29]. Although these two criteria have some similarities, they mainly reflect the conflicting approaches regarding the definition and clinical features of cervicogenic headache discussed above (see 'Definition and clinical features' above).
The CHISG criteria has an obligatory requirement for symptoms or signs of neck involvement that may be fulfilled if head pain is precipitated by neck movement, sustained awkward head positioning, or by external pressure over the upper cervical or occipital region on the symptomatic side [22]. This criterion may also be fulfilled by the combination of both restricted neck range of motion and ipsilateral neck, shoulder, or arm pain that is vague and nonradicular or radicular in nature. The CHISG criteria also require confirmatory evidence by anesthetic blocks, but this requirement is only obligatory for "scientific works."
The CHISG criteria describe the headache as nonthrobbing and nonlancinating, but these are not obligatory features. In addition, the CHISG criteria permit the occasional presence of certain migrainous features, including nausea, phonophobia and photophobia, dizziness, and ipsilateral blurred vision.
The ICHD-3 criteria for cervicogenic headache are as follows [29]:
●A) Any headache fulfilling criterion C
●B) Clinical and/or imaging evidence of a disorder or lesion within the cervical spine or soft tissues of the neck, known to be able to cause headache
●C) Evidence of causation demonstrated by at least two of the following:
•Headache has developed in temporal relation to the onset of the cervical disorder or appearance of the lesion
•Headache has significantly improved or resolved in parallel with improvement in or resolution of the cervical disorder or lesion
•Cervical range of motion is reduced and headache is made significantly worse by provocative maneuvers
•Headache is abolished following diagnostic blockade of a cervical structure or its nerve supply
●D) Not better accounted for by another ICHD-3 diagnosis
The ICHD-3 notes that while tumors, fractures, infections, and rheumatoid arthritis of the upper cervical spine have not been validated formally as causes of headache, these conditions are accepted as causes when demonstrated to be so in individual cases [29]. Similarly, cervical spondylosis and osteochondritis may or may not be valid causes of cervicogenic headache, depending upon the individual case.
Importantly, the ICHD-3 criteria state that the headache should be coded as tension-type headache (see "Tension-type headache in adults: Pathophysiology, clinical features, and diagnosis") when cervical myofascial pain is believed to be the cause of headache. This distinction emphasizes the point that cervicogenic headache requires dysfunction of cervical musculoskeletal elements (zygapophyseal joints) or upper cervical nerves innervating them and, in the strictest sense, is not considered a myofascial phenomenon. "Headache attributed to cervical myofascial pain" is listed only as an appendix diagnosis in the ICHD-3, indicating that it awaits further evidence and study to determine if it is more closely related to cervicogenic headache than to tension-type headache and before it can be considered a distinct entity [29]. Its inclusion in the ICHD-3 appendix as a potential diagnosis in need of further study is an acknowledgement of the frequent overlap of these two categories and challenge in differentiating.
The ICHD-3 appendix also acknowledges that upper cervical radiculopathy, which is rare, has been postulated as a cause of headache pending further evidence [29]. This is a rationale conclusion based upon the convergence of trigeminal and upper cervical sensory input at the trigeminocervical nucleus.
The utility of the CHISG criteria for cervicogenic headache appear to be limited by the nonspecific nature of the clinical features. A critical review of cervicogenic headache diagnostic criteria concluded that although neck structures play a role in the pathophysiology of some headaches, clinical patterns indicating a neck-headache relationship have not been adequately defined [4].
The ICHD-3 criteria acknowledge this shortcoming, noting that a number of features associated with cervicogenic headache are not unique to it [29]. These include focal neck tenderness, mechanical exacerbation of pain, unilaterality, posterior-to-anterior radiation of pain, nausea, vomiting, and photophobia. These features of cervicogenic headache do not define the relationship between the disorder and the source of the headache.
Diagnostic anesthetic blockade — Diagnostic anesthetic blocks for cervicogenic headache are not generally performed as office procedures and are not readily available to many practitioners [25]. Rather, these anesthetic blocks require specialized skills and facilities, a limitation that adds to the difficulty of confirming the diagnosis.
Diagnostic anesthetic blockade for the evaluation of cervicogenic headache are directed to several anatomic structures [6,9,30]:
●The lateral atlanto-axial joint, by intraarticular blocks
●The C2-3 zygapophyseal joint by intraarticular blocks or preferentially by blocking the medial branches of the C2 and C3 dorsal rami and the third occipital nerve where it crosses the joint
●The C3-4 zygapophyseal joint, by intraarticular blocks or preferentially by blocking the medial branches of the C3 and C4 dorsal rami
●Mid-cervical structures (the C4-5 and C5-6 zygapophyseal joints) more commonly refer pain to the neck and shoulders than to head but can refer to the occipital region and could be considered for diagnostic blockade with posterior predominant headaches. Patients whose headaches emanate from mid-cervical structures will not have frontal or orbital pain, as these mid-cervical structures do not have sensory input into the trigeminocervical nucleus, unlike upper cervical pain generators. This population is therefore often excluded from strict definitions of cervicogenic headache but remains important to consider clinically [23].
When the diagnosis is in doubt, anesthetic blocks should be performed in a blinded fashion using saline or other appropriate controls in order to identify true-positive cases and exclude placebo responders [31,32]. Because of ethical questions with performing placebo saline injections, differential local anesthetic blocks are often utilized in clinical practice to differentiate false-positive (ie, placebo) responses from true-positive responses. Two diagnostic blocks of the suspected pain generator are performed at different times, using a short acting anesthetic for one block and a long acting local anesthetic for the next. The patient is blinded to the anesthetic used. The response is interpreted by both the degree of improvement and the duration of improvement, which should be concordant with the local anesthetic used.
Imaging-guided blockade (using fluoroscopy, CT, or ultrasound) is necessary to assure accurate and specific localization of the pain source [33-35].
Importantly occipital neuralgia is classified as a distinct entity (see "Occipital neuralgia") and is not considered a cervicogenic headache [29]. The greater occipital nerve is a peripheral nerve extension of the medial branch of the dorsal ramus of the C2 cervical spinal nerve. Occipital neuralgia pathogenically stems from impingement or dysfunction of the greater occipital nerve as it courses through the soft tissues of the posterior neck and across the nuchal ridge. Accordingly, there is no anatomic rationale for greater occipital nerve blockade in the diagnosis or treatment of cervicogenic headache [6]. Such blockade is performed close to where the nerve crosses the superior nuchal line or at the level of C2 in the plane above the inferior obliquus capitis muscle (with ultrasound guidance), beyond which the nerve supplies the scalp of the occipital region. Thus, greater occipital nerve blockade does not confirm a cervical source of pain.
DIFFERENTIAL DIAGNOSIS — A patient with cervicogenic headache should have head and neck pain that are clearly interrelated to one another. Because there is a high prevalence of primary headache disorders such as migraine (episodic or chronic) and tension-type headache and a high prevalence of mechanical axial neck pain from cervical spondylosis or myofascial pain generators, they may exist as comorbid conditions in the same patient. In this situation, the head pain syndrome and the neck pain wax and wane independent of one another. Although this can be difficult to discriminate by history, it is an important point to explore.
A number of conditions in addition to cervicogenic headache can present with interrelated neck and head pain. Internal carotid or vertebral artery dissection, which present with neck pain and headache, is an important consideration in the differential diagnosis of acute cervicogenic headache [36-38]. In addition, the differential includes posterior fossa tumor, Arnold-Chiari malformation, herniated intervertebral disc, spinal nerve compression or tumor, arteriovenous malformation, and intramedullary or extramedullary spinal tumor.
Migraine and tension-type headache — Neck pain and muscle tension are common symptoms of a migraine attack [1,39-41]. Muscle tenderness in the posterior head and upper neck is common in tension-type headache as well. (See "Pathophysiology, clinical manifestations, and diagnosis of migraine in adults" and "Tension-type headache in adults: Pathophysiology, clinical features, and diagnosis".)
In one study of 50 patients with migraine, neck pain or stiffness associated with the migraine attack was reported by 32 patients (64 percent) [1]. Referred pain into the ipsilateral shoulder was present in seven (14 percent). In another report of 144 patients, neck pain associated with migraine attacks was noted by 108 patients (75 percent) [40]. Of these 108 patients, the pain was described as tightness by 69 percent, stiffness by 17 percent, and throbbing by 5 percent. The neck pain was unilateral in 57 percent, among whom it was ipsilateral to the side of headache in 98 percent.
C2 spinal nerve lesions — There are two distinctive clinical syndromes caused by lesions or injury affecting the C2 spinal nerve that may be confused with cervicogenic headache [6]:
●C2 neuralgia is characterized by a paroxysmal sharp or shock-like pain centered in the occipital region [42-44]. Ipsilateral eye lacrimation and conjunctival injection are common associated signs. These features contrast with the dull aching pain associated with cervicogenic headache [6]. Arterial or venous compression of the C2 spinal nerve or its dorsal root ganglion has been implicated as a cause for C2 neuralgia in some cases [42,45-47]. Meningioma has been rarely implicated [48].
●Neck-tongue syndrome is precipitated when rapid head turning causes subluxation of the posterior atlanto-axial joint and C2 spinal root compression [49-51]. The symptoms include neck pain, occipital pain, or both, often associated with ipsilateral tongue sensory symptoms. Onset is typically during childhood or adolescence.
Occipital neuralgia — Occipital neuralgia is a pain disorder characterized by paroxysmal jabbing pain in the distribution of the greater or lesser occipital nerves or of the third occipital nerve, sometimes accompanied by diminished sensation or dysesthesia in the affected area. In the classic description of occipital neuralgia, additional features include constant deep or burning pain with superimposed paroxysms of shooting pain. The pain is temporarily reduced or abolished by anesthetic blockade of the nerve. (See "Occipital neuralgia".)
Occipital neuralgia must be distinguished from occipital referral of pain from the atlanto-axial or upper zygapophyseal joints or from tender trigger points in neck muscles or their insertions. In our experience, this distinction may be difficult despite careful evaluation of clinical features and thorough examination. In such cases, the use of image-directed anesthetic blockade can usually permit an accurate diagnosis.
Occipital neuralgia is discussed in detail elsewhere. (See "Occipital neuralgia".)
TREATMENT — There is no proven effective treatment for cervicogenic headache. However, a number of different treatment modalities are available. Physical therapy is the preferred initial treatment because it is noninvasive. The available evidence suggests that pharmacologic therapy and botulinum toxin injections are not beneficial [6,52].
Physical therapy — Physical therapy may provide long-term improvement for cervicogenic headache, but the conclusions of different systematic reviews have differed about the strength of the effect size and the quality of the evidence:
●Based upon six randomized controlled trials that investigated cervicogenic headache [53-58], a 2013 systematic review concluded that there is good quality evidence supporting the use of cervical manipulation and mobilization, along with exercise, for the treatment of cervicogenic headache [59]. In the largest of the six included trials, which evaluated 200 patients with cervicogenic headache, treatment was unblinded and assessment was blinded [53]. Compared with controls who received no treatment, those assigned to six weeks of active treatment with either manipulative therapy, low-load endurance exercise therapy, or a combination of both therapies had a significant reduction in headache frequency at 12 months. The effect size was reported as moderate and clinically relevant. Combined treatment was not significantly better than either manipulative therapy or exercise alone.
Of concern, the cervical joint manipulation regimen used in some of the included trials consisted of both low-velocity and high-velocity techniques [53,55,58]. The concern arises because high-velocity chiropractic manipulation has been linked to a risk of arterial dissection and stroke in the vertebrobasilar territory. This issue is discussed separately. (See "Spinal manipulation in the treatment of musculoskeletal pain", section on 'Serious adverse events'.)
●Based upon four trials of participants with cervicogenic headache that compared physical therapy interventions with a control group or usual care [53,54,60,61], a 2016 systematic review and meta-analysis found that physical therapy led to a statistically significant benefit for reduced pain, frequency, and duration of cervicogenic headache [62]. However, the benefits did not necessarily achieve a clinically important effect size, and the quality of the evidence was considered low due to methodologic limitations of the trials.
●An open-label, two-site, controlled trial of spinal manipulation therapy (SMT) randomly assigned 256 patients with chronic cervicogenic headache to 0, 6, 12, or 18 SMT visits. Patients received a brief, light massage as control if not receiving SMT. The number of cervicogenic headache days was reduced in all groups (by one-third in the control massage group and by one-half in the SMT group); there was a linear dose-response with a reduction of one headache day/four weeks for every additional six SMT sessions. These responses were sustained to the last assessment at 52 weeks. No treatment-related serious adverse events were reported [63].
●A controlled trial enrolled 65 adults (50 to 75 years of age) with recurrent headaches associated with neck pain and musculoskeletal dysfunction, irrespective of headache classification [64]. The patients were randomly assigned to 14 physiotherapy sessions over 10 weeks or to usual care; physiotherapy sessions included low-velocity cervical mobilization and low-load therapeutic exercises of cervical flexors, axioscapular muscles, and postural correction exercises. Those receiving physiotherapy reported a reduction in headache frequency at the end of treatment (-1.6 days per week) and at six-month follow-up (-1.7 days per week) compared with usual care.
Despite the shortcomings of these data and modest to moderate effect size, we suggest physical therapy with exercise as the initial treatment option for patients with cervicogenic headache. We recommend not using high-velocity manipulation therapy because it is associated with a risk of stroke, and because the available limited evidence suggests that manipulation has no advantage over exercise therapy.
In our clinical experience, the intensity of headache may initially worsen during or after physical therapy, especially if it is vigorously applied. Physical treatment is better tolerated when initiated with gentle muscle stretching and manual cervical traction. Therapy can be slowly advanced as tolerated to include strengthening and aerobic conditioning. Using anesthetic blockade for temporary pain relief may enhance patient tolerance of physical therapy.
Pharmacologic treatments — Pregabalin was found to be effective in a small (n = 41) randomized controlled trial of patients with cervicogenic headache in decreasing the average number of headache days per month significantly in the pregabalin group (10 fewer headache days per month compared with no change in headache days per month in the placebo cohort [65]. This benefit was evident at the end of a 12-week treatment trial, but was not seen at the end of the first four weeks of treatment during which time patients were titrating the dosage up. Most patients were treated with 450 mg per day (split twice a day), but dosing was variable and based on patient tolerance.
While duloxetine has not been studied specifically for cervicogenic headache, it is US Food and Drug Administration-approved for chronic musculoskeletal pain. Cervicogenic headache by definition is generated by several axial spine musculoskeletal joints.
Common medications used for headache disorders (eg, tricyclic antidepressants, gabapentin) are often and reasonably trialed in patients with cervicogenic headache, but they have not been systemically studied in this patient population.
Anesthetic blockade — Anesthetic injections of the lateral atlanto-axial joint, the C2-3 zygapophyseal joint (and the overlying third occipital nerve), and/or the C3-4 zygapophyseal joint can temporarily reduce or relieve pain and may allow greater participation in physical treatments.
Greater and lesser occipital nerve blockade may provide temporary pain relief in some cases, but the benefit of this treatment is not specific for cervicogenic headache [6,66]. In addition, there is no anatomic rationale for greater occipital nerve blockade in the diagnosis or treatment of cervicogenic headache. (See 'Diagnostic anesthetic blockade' above.)
Radiofrequency neurotomy — Radiofrequency techniques may offer potential benefit for cervicogenic headache, but that benefit has not been confirmed in adequate randomized controlled trials with sufficient sample size. Recognizing the limitations of available data, we suggest radiofrequency neurotomy for patients with cervicogenic headache arising from the zygapophyseal joints who have the diagnosis confirmed by near complete (>90 percent) pain relief in response to controlled anesthetic block and who are refractory to noninvasive therapy. The benefit of radiofrequency neurotomy for other etiologies of cervicogenic headache is unknown.
Percutaneous radiofrequency neurotomy can be considered for cervicogenic headache arising from the C2-3 or C3-4 zygapophyseal joint if diagnostic anesthetic nerve blockade is temporarily successful in providing near-complete pain relief. However, the available evidence is limited and conflicting.
A 2018 systematic review of radiofrequency ablation neurectomy and pulsed radiofrequency therapy identified 10 trials that met inclusion criteria with the following observations [67]:
●There were three small randomized controlled trials [68-70]. Two trials (total 42 patients) found no significant benefit [68,69]; in one of these, patients were recruited by clinical criteria rather than by gold-standard diagnostic blocks [68]. One trial (n = 30) randomly assigned patients to greater occipital nerve block with glucocorticoid injection or to pulsed radiofrequency ablation; there was improvement in pain levels in both groups at three months, but only the pulsed radiofrequency ablation cohort showed sustained improvement at nine months [70].
●There were seven non-randomized studies. Three were small prospective studies, which found that radiofrequency ablation neurectomy was associated with improvement in cervicogenic headache [69,71-73]. Three were retrospective observational studies of radiofrequency ablation or pulsed radiofrequency therapy of the C1-2 zygapophyseal joint, C2 dorsal root ganglion, and/or the third occipital nerve [74-76], and one was a retrospective observational study of radiofrequency ablation of lower cervical medial branches (C4-7) [15]; these all reported that treatment was associated with improvement in cervicogenic headache.
Glucocorticoid injection — Randomized controlled trials evaluating glucocorticoid injections for cervicogenic headache are lacking. However, results from small retrospective studies suggest that some patients may obtain relief from intraarticular glucocorticoid injections [77-79].
●In one report, 18 patients with cervicogenic headache had fluoroscopically guided therapeutic intraarticular glucocorticoid injections at the C2-3 zygapophyseal joint. At 19 months after injection, reduced headache frequency was noted for 11 patients (61 percent), and freedom from pain for two (11 percent).
●Another study identified 32 patients cervicogenic headache related to atlanto-axial joint pain, confirmed by diagnostic block [78]. Immediately following intraarticular glucocorticoid injections, a 50 percent or greater decrease in pain scores was noted by 26 patients (82 percent). However, at six months, the difference between pre- and postprocedure pain scores was not statistically significant.
●A third report included 31 patients who were treated with standardized injections of local anesthetic and glucocorticoid directed to multiple potential cervicogenic headache pain generators (C1/2 and C2/3 facet joint injections, the C2 dorsal ramus, and the C3 dorsal and ventral rami) in one procedural appointment [79]. Treatment was associated with a >50 percent headache relief in 90 percent of patients, sustained for an average duration of 22 days [79].
Coblation — Coblation is a bipolar technique that passes radiofrequency energy at low temperatures through a conductive medium to produce a plasma field causing tissue destruction but is not commonly used as a pain intervention. One retrospective, single-center study treated 26 patients with cervicogenic headache of at least moderate severity with ultrasound guided C2 nerve root coblation [80]. At 24 weeks of follow-up, treatment was associated with a decrease in pain scores by ≥50 percent in 92 percent of patients. This technique requires further study.
Neuromodulation — Electrical stimulation pain modulation through spinal cord and peripheral nerve stimulation (ie, neuromodulation) has made tremendous technological advances over the last decade and is now widely used in the treatment of refractory neuropathic pain conditions. A three-year retrospective study evaluated occipital nerve stimulation for 16 patients with refractory cervicogenic headache who were implanted between 2011 and 2013 [81]. At one year, treatment was associated with a 50 percent or more improvement in headache pain scores for 11 patients (69 percent); at three years post-implant, a similar level of improvement was noted for 6 patients (38 percent) [81].
Further validation of this technique is necessary in cervicogenic headache, but it is notable that neuromodulation stimulation options, including stimulation frequency and burst stimulation programming, have advanced substantially since the 2011 to 2013 implant period of the cited study [81]. In addition, new peripheral nerve stimulation techniques offer less invasive and burdensome opportunities to stimulate the occipital nerves compared with traditional neuromodulation techniques available at the time of that study. Neuromodulation allows the ability to do test stimulation to judge efficacy prior to an outpatient surgical implantation.
Surgery — A variety of surgical interventions have been performed for presumed cases of cervicogenic headache, but surgical procedures for cervicogenic headache are not recommended unless there is compelling evidence of a surgically amenable lesion on imaging causing the cervicogenic headache (ie, confirmed to be the pain generator by local anesthetic diagnostic blocks) that is refractory to all reasonable nonsurgical treatments.
The available data are limited to small retrospective studies, but suggest that surgery may be beneficial for three specific causes of cervicogenic headache: 1) C2 spinal nerve compression by vascular/ligamentous structures, 2) osteoarthritis of the lateral atlanto-axial joint, and 3) upper cervical intervertebral disc pathology.
●Surgical decompression and microsurgical neurolysis of the C2 spinal nerve were reported in one series of 31 patients who met clinical criteria for cervicogenic headache, including headache relief from diagnostic block of the C2 spinal nerve. Treatment was associated with complete pain relief in 14 patients (45 percent) and "adequate improvement" in 16 (52 percent) [46].
●Arthrodesis is an option for the treatment of cervical pain and headache related to osteoarthritis of the lateral atlanto-axial joint. In three studies that reported a total of 24 patients, complete relief of pain or an "excellent result" was achieved in almost all patients, with benefit lasting two or more years [82-84].
Intensification of pain and anesthesia dolorosa are potential adverse outcomes that must be considered when contemplating the use of surgical interventions.
INFORMATION FOR PATIENTS — UpToDate offers two types of patient education materials, "The Basics" and "Beyond the Basics." The Basics patient education pieces are written in plain language, at the 5th to 6th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10th to 12th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.
Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on "patient info" and the keyword(s) of interest.)
●Basics topics (see "Patient education: Neck pain (The Basics)" and "Patient education: Whiplash (The Basics)")
●Beyond the Basics topic (see "Patient education: Neck pain (Beyond the Basics)")
SUMMARY AND RECOMMENDATIONS
●The underlying pathophysiology and source of pain in cervicogenic headache are controversial. The most persuasive explanation is that cervicogenic headache is caused by referred pain from the upper cervical joints (figure 1). The first three cervical spinal nerves and their rami are the primary peripheral nerve structures that can refer pain to the head (figure 2). (See 'Etiology and anatomy' above.)
●There is no consensus regarding the definition of cervicogenic headache, but two conflicting viewpoints are widely recognized. One relies on clinical features, while the other ignores clinical features and relies on the demonstration of a distinct anatomic source in the neck that can cause head pain. The definitive criterion for the latter definition is near-complete relief of pain after controlled anesthetic blocks of cervical structures or their nerve supply. Conflicting diagnostic criteria have also been proposed. (See 'Definition and clinical features' above and 'Diagnosis' above.)
●Since accurate diagnosis of cervicogenic headache is a prerequisite to treatment, we suggest that the diagnosis should be confirmed by controlled anesthetic blocks of cervical structures or their nerve supply if available. Other headache etiologies should be pursued if the response to diagnostic blockade is incomplete. (See 'Diagnosis' above and 'Diagnostic anesthetic blockade' above.)
●Dissection of the internal carotid or vertebral artery is an important consideration in cases of suspected acute cervicogenic headache. In addition, the differential diagnosis includes posterior fossa lesions, migraine, and tension-type headache, particularly when head pain is chronic. (See 'Differential diagnosis' above.)
●There is no proven treatment for cervicogenic headache. For patients with probable cervicogenic headache that is diagnosed by clinical criteria, with or without confirmatory diagnostic anesthetic block, we suggest physical therapy as the initial treatment option (Grade 2C). However, high-velocity cervical manipulation techniques are associated with a risk of ischemic vertebrobasilar territory stroke and should not be used as part of the treatment regimen. (See 'Treatment' above and 'Physical therapy' above.)
●For patients with cervicogenic headache who are refractory to physical therapy and who have a source of pain confirmed by diagnostic blockade, treatment options depend on the site causing the headache:
•For patients with cervicogenic headache stemming from the C2-3 or C3-4 zygapophyseal joint, confirmed by complete pain relief with diagnostic blockade, we suggest radiofrequency neurotomy of the third occipital nerve and the medial branches of the C2, C3, and/or C4 dorsal rami (Grade 2C). For patients who do not want neurotomy, we suggest intraarticular glucocorticoid injections (Grade 2C). (See 'Radiofrequency neurotomy' above and 'Glucocorticoid injection' above.)
•For patients with cervicogenic headache stemming from the atlanto-axial joint, confirmed by complete pain relief with diagnostic blockade, who are refractory to physical therapy, we suggest arthrodesis of the joint (Grade 2C). (See 'Surgery' above.)
ACKNOWLEDGMENT — The editorial staff at UpToDate acknowledge Zahid H Bajwa, MD, who contributed to an earlier version of this topic review.
