Doi:10.1016/j.pain.2006.01.005

authors stated that there was no scalp numbness afterthey injected their solutions which contained Xylocaı¨neÒ With great interest we have read the article: ÔSuboccip- 2% 0.5 ml. This could indicate that the injection site was ital injection with a mixture of rapid- and long-acting not close enough to the greater occipital nerve or they steroids in cluster headache: A double-blind placebo-con- used a too small amount of local anaesthetic. A larger trolled studyÕ by Anna Ambrosini et al. in Pain amount of local anaesthetic, resulting in scalp anaesthe- sia, can confirm that their injection is close enough to In their article, the authors suggested an injection the greater occipital nerve. Besides, they could have used with a mixture of rapid- and long-acting betamethasone a nerve stimulator to locate the greater occipital nerve. Be- near the ipsilateral greater occipital nerve for the treat- cause we question the right side of injection used in this ment of cluster headache. In their study they included study, the only conclusion to be drawn from their study patients with a known cluster headache period of 4 is that an occipital subcutaneous injection of rapid- and weeks or longer. Based on their results they concluded long-acting betamethasone can prevent cluster headache that suboccipital corticoid injections near the greater attacks at the start of a cluster headache period.
occipital nerve significantly reduced the cluster headache Secondly, we have some concerns about the method- period compared to the normal expected length of the ology of this study. To analyse the causality of the great- er occipital nerve in the pathophysiology of cluster Despite this remarkable efficacy of their therapeutic headache, an additional patient group should have been approach in cluster headache patients, we would like included where patients received a subcutaneous injec- to comment on two aspects of their study: tion with betamethasone in another part of the body: At first we like to comment on the anatomical site of this would allow to make a differentiation with a system- the injection they used in their study. The authors inject- ic therapeutic effect of the corticoids. Another striking ed their solution halfway between the inion and the mas- feature was that 6 of the 11 positive responders to the toid (3–4 cm below the inion) to reach the greater injection with betamethasone and Xylocaı¨neÒ immedi- occipital nerve. The authors also stated that there is ately became attack free. Of these 6 immediate respond- no standard recommendation for the precise location ers 4 had a long-lasting effect. Since no patient in the of the suboccipital injection. Vital et al. (1989) studied placebo group had an immediate effect and since 4 pa- the course of 18 greater occipital nerves in 9 formalin tients had a long-lasting effect, the efficacy of their ther- embalmed adult cadavers (5 women and 4 men) and apy neither can be attributed to the use of local found that the greater occipital nerve crossed the trape- anaesthetic nor to betamethasone since the observed ef- zius muscle (where the nerve becomes subcutaneous) on fect occurred too quick after the injection. In our opin- average 31.8 mm from the midline and 22.2 mm below the external occipital protuberance. Ashkenazi and Lev-in (2004) also give an indication for the injection site of 1. These patients were placebo responders although we the greater occipital nerve. They locate the subcutaneous admit it is odd that then there were no placebo part of the greater occipital nerve 3.5 cm inferolaterally to the occipital protuberance. This position approxi- 2. The betamethasone and Xylocaı¨neÒ were injected mates the description by Vital JM et al. The spot where the greater occipital nerve becomes subcutaneous as de-scribed in Staubesand (1988) is not halfway between ini- To prove the role of the greater occipital nerve as part on and mastoid but closer to the inion. Brown (1996) of the pathophysiology of cluster headache, it is our also locates the injection point closer to the inion. Based opinion that more randomised clinical trials are neces- on these literature data we wonder whether the site of sary where on one hand the proximity of the injection injection chosen by the authors is located near the great- site near the greater occipital nerve needs to be con- er occipital nerve. This is supported by the fact that the firmed and where on the other hand an extra control Letters to the Editor / Pain 121 (2006) 281–284 group should be included where corticoids are injected Vanelderen, but at a deeper site where the nerve is in another part of the body to rule out a systemic corti- more lateral (see SobottaÕs Atlas). We agree that coid effect. Due to the concerns expressed above we the injection did not produce numbness in most sub- wonder if this study permits us to conclude that an injec- jects probably because of the small amount of xylo- tion with betamethasone near the greater occipital nerve caine (0.5 ml) and the fact that the GON was not is a valuable technique to treat cluster headache.
2. Vanelderen et al. underscore a potential shortcom- ing of our study which we have discussed properly in the discussion section, i.e., is the therapeuticbenefit mediated by a systemic effect of the steroid, Vital JM, Grenier F, Dautheribes M, Baspeyre H, Lavignolle B, and thus would a systemic injection be equally Se´ne´gas J. An anatomic and dynamic study of the greater occipitalnerve (n. of Arnold). Surg Radiol Anat 1989;11:205–10.
beneficial? As mentioned in the article, we hypoth- Ashkenazi A, Levin M. Three common neuralgias. Postgrad Med esise that the suboccipital site of injection is of importance for 2 reasons: first, for systemic admin- Brown CR. Occipital neuralgia: symptoms, diagnosis and treatment.
istrations much higher doses of steroids are neces- Pract Periodontics Aesthet Dent 1996;8:587–8.
sary to obtain an effect (Cianchetti et al., 1998; Staubesand J. Sobotta Atlas of Human Anatomy. Munich – Vienna – Baltimore: Urban and Schwarzenberg; 1988.
Mir et al., 2003) and, second, intramuscular injec-tions of 120 mg prednisolone were found ineffectivein one study (Anthony, 1987). However, we also state in the discussion section that the final proof has to come from a trial comparing suboccipital Critical Care and Multidisciplinary Pain Centre, and sytemic steroid injections, which we are plan- ning. Although initially considered, we decided E-mail address: [email protected] present study for logistic reasons and becauserecruitment of sufficient patients would have been 3. The immediate therapeutic response in the verum arm suggests, according to Vanelderen et al., that the patients were placebo responders or that the injection was intravascular. As a matter of fact (see results table), the response was immediate (no attacks afterthe injection) only in 4 out of 8 patients with asustained response at 4 weeks and there was no * Corresponding author. Tel.: +32 496 087 926.
correlation with the duration of longer-lasting remis- 0304-3959/$32.00 Ó 2006 International Association for the Study ofPain. Published by Elsevier B.V. All rights reserved.
sion. Nevertheless, it is clinically well documented (Anthony, 1987) that the attacks may disappear imme-diately after the injection in a subgroup of patients. Asmentioned by Vanelderen et al. themselves this is unlikely to be a placebo effect, because it did not occurin the placebo arm. There is thus no need to speculate We thank Drs. Vanelderen et al for their interest in our on an intravascular injection, the more so that this study of suboccipital steroid injections in the treatment of was excluded by proper aspiration before injecting cluster headache. They criticise our study for the follow- the solutions and would have produced xylocaine-re- lated side effects at least in some patients.
1. The injection site may not have reached the greater To conclude, we think that our injections were in the occipital nerve. This may be possible and we do not vicinity of the deeper, more proximal, portion of the claim that the injection was hitting the GON, but greater suboccipital nerve and that our study demon- that it was in the vicinity of the nerve. As mentioned strates for the first time in a blinded placebo-controlled in the methods section, we purposely performed protocol that suboccipital steroid injections are indeed deep injections in close contact with the occipital an effective treatment for epsiodic and chronic cluster periosteum to avoid a known adverse effect of ste- headache, as claimed since a long time on empirical roid injections, local alopecia (Shields et al., 2004).
grounds. We agree that another comparative trial is nec- These injections do not target the GON at the site essary to demonstrate that systemic steroid injections where it becomes subcutaneous as argued by

Source: http://www.clusterheadaches.com/michael/2006_comment.pdf