Abstract
Objectives: A complete response to indomethacin is required for the diagnosis of hemicrania continua (HC). Nevertheless, patients may develop side effects leading to withdrawal of this drug. Several alternatives have been proposed with no consistent effectiveness. Both anaesthetic blocks of peripheral nerves and trochlear corticosteroid injections have been effective in some case reports.
Methods: Twenty-two patients with HC were assessed in a headache outpatient office. Physical examination included palpation of the supraorbital nerve (SON) and greater occipital nerve (GON) as well as of the trochlear area.
Results: In 14 patients, at least one tender point was detected. Due to indomethacin intolerance, at least one anaesthetic block of the GON or SON, or an injection of corticosteroids in the trochlear area, were performed in nine patients. Four of them were treated with a combination procedure. All these patients experienced total or partial improvement lasting from 2 to 10 months.
Conclusion: Anaesthetic blocks or corticosteroid injections may be effective in HC patients showing tenderness of the SON, GON or trochlear area.
Objectives
Characteristics of nine patients treated with anaesthetic block of SON or GON or trochlear corticosteroid injection
M: male; F: female; SON: supraorbital nerve; GON: greater occipital nerve.
Increasing attention has been recently focused on the influence of peripheral factors in the genesis or modulation of primary headaches. Anaesthetic blocks of peripheral nerves and trochlear corticosteroid injections have already shown effectiveness in some cases of HC (3–6).
The aim of this study was to evaluate the therapeutic contribution of peripheral nerve blocks or trochlear corticosteroid injections, based on the presence of nociceptive inputs, in a series of 22 patients with HC.
Methods
We prospectively assessed all consecutive new patients with HC attending a headache outpatient office in a tertiary hospital over a three-year period (January 2008–January 2011). We obtained a complete medical history from each patient, including age at headache onset. The characteristics of both background pain (side, site, type, intensity) and pain exacerbations (frequency, intensity, periodicity, associated autonomic symptoms) were carefully recorded. Subsequently, we performed a complete physical and neurological examination, including palpation of the main superficial arteries and pericranial nerves. The supraorbital nerve (SON), minor occipital nerve, greater occipital nerve (GON) and the trochlear area were palpated systematically in search for local tenderness. Magnetic resonance imaging (MRI) or computerized tomography (CT) of the head was carried out in all patients.
We assessed indomethacin response with a standard oral trial, up to 250 mg per day (1,7). To confirm HC diagnosis, we took special care to assure complete response to indomethacin in all patients before any discontinuation of therapy due to side effects.
Results
Twenty-two patients (14 females, eight males), out of the 1150 who attended in the aforementioned outpatient office during the inclusion period (1.9%), were diagnosed with HC. Mean age at onset was 41.8 ± 18.1 years (range: 7–74 years). Pain was strictly right-sided in 14 patients (63.6%) and left-sided in eight (36.4%).
Patients rated pain intensity as 5.2 ± 1.2 for the background pain and 8.4 ± 1.1 for the exacerbations on a verbal analogue scale (VAS) (0: no pain, 10: the worst imaginable pain). Three patients (13.6%) suffered episodic instead of continuous pain. All patients presented exacerbations and five (22.7%) did not report associated autonomic symptoms.
In 14 patients (63.6%) the neurological examination revealed local tenderness ipsilateral to the pain in at least one point among the SON, GON or trochlear area. Nine patients did not tolerate indomethacin and were treated with at least one anaesthetic block of the GON or SON or a trochlear corticosteroid injection. In four of these patients a combination procedure was performed (SON + GON in patients 4, 5 and 9; SON and trochlear area in patient 3), as shown in Table 1. We blocked the GON by injecting 2 cc of 0.5% bupivacaine and 2% mepivacaine in a 1:1 ratio, at one-third of the distance between the external occipital protuberance and the mastoid process (8). We blocked the SON by injecting 0.5 cc of the same mixture 2 cm lateral to the medial border of the eyebrow, next to the supraorbital notch (8). In patients with trochlear tenderness, we injected 4 mg of triamcinolone into the symptomatic region (6). We did not use steroids when blocking the GON or SON to avoid the trophic changes occasionally described at those sites, which may be especially disfiguring at the supraorbital area.
Five patients reported complete relief with nerve block comparable to their pain-free period with indomethacin. The remaining four patients showed partial improvement in pain intensity as measured by the VAS (at least three points in both background pain and exacerbations); this partial response was similar among these four patients. Pain relief began immediately after the procedure and lasted from 2 to 10 months. When blocks were repeated, response length increased with subsequent procedures. The main characteristics of these nine patients are summarized in Table 1.
Discussion
The requirement of complete response to indomethacin for HC diagnosis has been a contentious issue (9,10). In our study, we have taken such response to indomethacin as a diagnostic criterion, just as in other recent studies (1). Indomethacin may have a different effect to other anti-inflammatory drugs by inhibiting nitric oxide-induced dilation of the dural meningeal vessels (11). It has been proposed that indomethacin might act as a disease-modifying agent, since some patients may reduce the dose, or even give up the medication, while remaining pain-free (12,13).
About three quarters of patients report side effects when receiving therapeutic doses of indomethacin. The most frequent side effects are abdominal symptoms, ranging from discomfort to gastrointestinal ulceration and bleeding, and dizziness. These complications may even lead to withdrawal of this drug in up to 20% of patients (1,13–15).
Several other drugs have been proposed as alternative therapies in HC (9), such as acemetacin, celecoxib (15), rofecoxib (15), piroxicam, acetylsalicylic acid, ibuprofen (14), caffeine, topiramate (1), lamotrigine (9), gabapentin, steroids, melatonin, verapamil (10), amitriptyline, lithium (9) and methysergide (1,9), but none of these has been found consistently effective.
This complex therapeutic puzzle, in which a well-defined first therapeutic choice is often not tolerated, and with no consistent alternatives, has brought some interest in the role of blocks of peripheral nociceptive inputs in the treatment of HC (6).
It is known that interventions on peripheral nerves might be useful in the treatment of primary headaches and they have been widely used, although there are no proven indications (8). Nerve blocks may relieve pain by reducing afferent transmission of nociceptive pathways, as a neural block in a sensitized nerve might wind down central sensitization. Although the local anaesthetic effect lasts no more than 6 h, it may be crucial in breaking the pain cycle which leads to chronic pain. Therefore, the benefit lasts longer than expected considering the short duration of the block (3,5,8). At least in migraine, no differences have been found when blocking nerves with anaesthetic alone, or with the addition of a steroid (16).
The GON is regarded as a possible target in the therapy of chronic headaches. In fact, GON blocks have proven effective for cervicogenic headache, cluster headache and occipital neuralgia. In addition, favourable results have been reported in migraine, medication-overuse headache and tension-type headache (3,8,17,18). Yet, there is no standard procedure for GON block, concerning the site of injection, the association of steroids or the use of tenderness as a selection criterion (3). Occipital tenderness has been observed in up to 67% of HC patients, and may be ipsilateral, or even bilateral, to the pain side (7). The GON may be considered a therapeutic target in HC, as occipital nerve stimulation has shown promising results in the treatment of this syndrome (19). Some authors have already used anaesthetic blocks in HC patients. In a recently published study, GON anaesthetic blocks were performed in 23 patients, with a reported effectiveness of 35% (1). Previously, in shorter series, GON blocks with anaesthetic alone (4), or with anaesthetic plus steroid (5), did not provide a significant benefit. However, local tenderness was not used as a selection criterion as it was in our study.
SON block is an effective treatment for supraorbital neuralgia, both spontaneous and posttraumatic (8), and one study has shown favourable response after SON block in migraine patients (20). Regarding HC, only one study has reported partial response of HC to anaesthetic block of the SON, when performed after a GON block (4).
Local tenderness in the trochlea is the main feature of primary trochlear headache (21) and is occasionally found in other primary headaches (6). A therapeutic benefit of trochlear corticosteroid injection in both HC and paroxysmal hemicrania has been described in isolated cases (6,22).
The patients in this study had a surprisingly long effect of these peripheral procedures. It is hard to explain why the response may be more long-lasting in HC than in other primary headaches. Perhaps this reflects an important role of external triggers or peripheral modulators in the pathophysiology of HC.
In conclusion, this study shows that both anaesthetic blocks of the GON or SON and trochlear steroid injections might be therapeutic alternatives in patients with HC who are intolerant of indomethacin. The small sample size and the non-controlled design are the limitations of this study. Placebo-controlled trials to evaluate the efficacy of these procedures in HC as well as other trigeminal autonomic cephalalgias should be conducted. We propose to begin with anaesthetics alone to avoid the risks associated with steroid injections. Additionally, we suggest considering local tenderness as a selection criterion in future trials.
Footnotes
Funding
This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.
Declaration of interest
The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper. This paper was partially presented as a Poster at the XIV Congress of the European Federation of Neurological Societies (EFNS), Geneva, Switzerland, 25–28 September 2010.