Abstract
Upper cervical pain is frequent in different primary headaches and not sufficient evidence for cervicogenic headache (CH). Biological markers should help to differentiate CH from other headache disorders. In most cases, imaging techniques of the cervical spine are not helpful for the diagnosis of CH. Symptoms and signs of neck involvement, such as a mechanical precipitation of attacks, a restriction in range of motion of the cervical spine, and the existence of ipsilateral neck, shoulder, or arm pain, seem to be reasonably valid for the diagnosis of CH, but its reliability and validity should be confirmed in larger studies. Positive diagnostic blockades of cervical structures or its nerve supply are not specific for CH. Neurophysiological investigations give some insight into the pathophysiological mechanisms of CH but are not diagnostic. In CH, calcitonin gene-related peptide levels do not differ between the symptomatic and the asymptomatic side, between the jugular and the cubital blood, and between days with and without headache. There is no evidence for an activation of the trigeminovascular system in CH. It can be concluded that CH is not just a migraine variant triggered by neck dysfunction but a functional entity.
Introduction
Almost 25 years after its first systematic description in 1983, the existence of cervicogenic headache (CH) is largely accepted (1). Nevertheless, there is a controversy about the quality of diagnostic criteria. Those diagnostic criteria provided by the Cervicogenic Headache International Study Group (CHISG) differ from the criteria provided by the International Headache Society (IHS), which are shown in Table 1 (2, 3). The crucial question is whether the diagnostic criteria sufficiently discriminate CH from other, especially from primary, headache disorders. There is a high frequency of upper cervical pain also in migraine, tension-type headache (TTH) and different trigemino-autonomic cephalgias; thus the existence of cervical pain is not sufficient for the diagnosis of CH. In the following, the value of potential biological markers (imaging techniques, clinical signs, diagnostic blockades, neurophysiological investigations, and laboratory testing) for the diagnosis of CH is discussed.
IHS diagnostic criteria for cervicogenic headache, diagnosis 11.2.1 (3)
Tumours, fractures, infections and rheumatoid arthritis of the upper cervical spine have not been validated formally as causes of headache, but are nevertheless accepted as valid causes when demonstrated to be so in individual cases. Cervical spondylosis and osteochondritis are NOT accepted as valid causes fulfilling criterion B. When myofacial tender spots are the cause, the headache should be coded under 2. Tension-type headache.
Clinical signs acceptable for criterion C1 must have demonstrated reliability and validity. The future task is the identification of such reliable and valid operational tests. Clinical features such as neck pain, focal neck tenderness, history of neck trauma, mechanical exacerbation of pain, unilaterality, coexisting shoulder pain, reduced range of motion in the neck, nuchal onset, nausea, vomiting, photophobia, etc., are not unique to cervicogenic headache. These may be features of cervicogenic headache, but they do not define the relationship between the disorder and the source of the headache.
Abolition of headache means complete relief of headache, indicated by a score of zero on a visual analogue scale (VAS). Nevertheless, acceptable as fulfilling criterion C2 is ≥ 90% reduction in pain to a level of < 5 on a 100-point VAS.
Imaging techniques
Typical CH is often not accompanied by radiological pathologies on standard X-ray. Conversely, even severe degenerative changes of the cervical spine do not necessarily evoke pain (4). Typically, no specific abnormalities are found on cervical MRI. There is a similar prevalence of disc bulging in CH as with healthy controls (5). Therefore, even modern imaging techniques such as MRI are of insufficient sensitivity and specificity for the diagnosis of CH.
Clinical signs
The IHS criteria demand evidence that the pain in CH can be attributed to a neck disorder. This evidence could be theoretically drawn from a demonstration of clinical signs that implicate a source of pain in the neck (criterion C1, Table 1). However, the IHS denies the reliability and validity of such clinical signs as defined by the CHISG in a note saying that symptoms and signs of neck involvement are not unique to CH and that they may be features of CH but do not define the source of pain. As a matter of fact, the validity of symptoms and signs of neck involvement for the diagnosis of CH has already been studied (6). Unilaterality without side shift, a mechanical precipitation of attacks, a restriction in range of motion of the cervical spine, and the existence of ipsilateral neck, shoulder or arm pain, were often encountered in CH but only infrequently in migraine and TTH. It was concluded that these clinical criteria adequately distinguish CH from migraine and TTH.
Diagnostic blockades
According to the IHS diagnostic criteria, evidence that the pain can be attributed to the neck can be alternatively based on positive diagnostic blockades using controlled techniques and achieving a complete or almost complete relief of pain (criterion C2, Table 1). Indeed, a blockade of the greater occipital nerve (GON), which is the most often routinely applied technique, demonstrated a high sensitivity in CH of approximately 80% positive responses (7). On the other hand, the specificity of diagnostic blockades is low according to several different studies reporting positive responses in more than 45% of patients suffering from primary headache disorders such as migraine and cluster headache (8, 9).
Neurophysiological studies
Most recently, the blink reflex has been investigated in patients with CH (10). As compared with migraine and TTH, there was a decreased amplitude of the R2 component and a reduced R2 duration after stimulating the symptomatic side in CH, pointing to an asymmetric brainstem excitability. This finding might give some insight into the pathophysiological mechanisms of CH. However, it was a group effect with an expanded overlap of values between the different headache groups, making the method worthless for diagnostic purposes.
Laboratory testing
From clinical experience, the most difficult differential diagnosis of CH is migraine without aura. One of the most important biological markers for migraine is an increase of calcitonin gene-related peptide (CGRP) in the cranial circulation during the attack, pointing to a phenomenon called trigeminovascular activation (11). Replicated by other groups and proven valid in a recent proof of concept study with a CGRP antagonist, CGRP release is an accepted biological marker for migraine attacks, albeit a recent Scandinavian study did not corroborate this hypothesis (12). In order to investigate a possible activation of the trigeminovascular system in CH, we identified 11 patients fulfilling all diagnostic criteria of the CHISG (13). CGRP measurements were performed with a well-established radio-immunoassay in blood from both external jugular veins and a cubital vein. There were no differences at all between the symptomatic and the asymptomatic side, between the jugular and the cubital blood, and between days with and without headaches. In conclusion, there is no trigeminovascular activation in CH, and this headache entity is biologically different from migraine.
Conclusions
Imaging techniques of the cervical spine are not helpful for the diagnosis of CH in the majority of cases. The so called ‘symptoms and signs of neck involvement’ as defined by the CHISG seem to be reasonably valid for the diagnosis of CH but its reliability and validity should be confirmed in larger studies. A positive diagnostic GON blockade is not specific for CH. Neurophysiological investigations give some insight into the pathophysiological mechanisms of CH but are not diagnostic. There is no evidence for trigeminovascular activation in CH, and it can be concluded that CH is not just a migraine variant triggered by neck dysfunction.
Conflicts of interest
SE and AF have declared no conflicts of interest.