Abstract
We are delighted to hear that our recently published study “High cervical spinal cord stimulation for chronic cluster headache” has raised the interest of renowned experts in this field.
The points of criticism pertain first to the definition of refractory cluster headache, thus to patient selection, and second to major adverse events, i.e. to the high rate of complications.
We assure you that in all of our patients the selection criteria as proposed by Leone et al. (1) were applied. As Dr Gaul and colleagues point out, only one of seven patients had verapamil in a dose above 480 mg/day, but in this context it should be also stated that all patients had trials of verapamil: two patients had 240 mg, three patients had 360 mg, one patient had 480 mg, and one patient had 640 mg/day. All of these doses were results of trials, in which higher doses had not been tolerated.
Likewise it was criticized that only two out of seven patients were treated with topiramat in doses >100 mg. However, all (except patient 1 who had received his first electrode in February 2004, prior to the publication of Leone et al. in November 2004 (1)) patients had trials of topiramat 100 mg/day and higher doses were not tolerated.
Maybe it should also be mentioned again that the mean baseline attack frequency in our study was slightly higher than in the two studies by Burns et al. (which describe a lesser clinical effect of occipital nerve stimulation (ONS) compared to high cervical spinal cord stimulation (hcSCS): 5.9 attacks/day vs. 3.9 and 4.8 attacks/day (2,3)) and much higher than in the study by Magis et al. (1.7 attacks/day (4)) which describes a somewhat similar effect strength compared to hcSCS.
We thank Dr Gaul and colleagues for giving us the possibility to clarify this point. The high incidence of lead complications, namely lead breakages certainly is a point with which we are highly incontent. We (in particular TW) are in a way pleased to note that Dr Gaul and colleagues attribute the complication rate to the method itself and not to a surgical “learning curve”. We think that probably both factors, those inherent to the method such as the high mobility of the cervical spine as well as those that can be corrected in the future, such as surgical technical details (electrode selection, lead fixation, use of paddle leads), may contribute to the complication rate. Moreover, we think that in both methods (ONS and SCS) to date, the numbers of patients treated of are still too small for definite conclusions on complication rate. Meanwhile we have treated one further patient with hcSCS without any complications. If also the next one or two patients should have no complications, the overall complication rate would probably resemble that of ONS.
Our aim was to present the results of this method to discussion and not to promote one or another surgical option to treat chronic cluster headache. Pros and cons of both methods should be carefully examined. In plain simple terms, pros and cons of ONS and hcSCS can be denominated as follows:
ONS
Pros: efficacy of different strength in most of the cases, easy implantation technique, less invasive Cons: in some patients limited efficacy, sometimes effect detectable only after several weeks, eventually leading to implantable pulse generator in patients who later turn out to be non-responders. hcSCS
Pros: direct mode of action, effect can be felt immediately (prophylactic effect within the first days), high average efficacy, to current knowledge better than some studies on ONS (2,3,5), comparable to other studies (4), efficacy in all patients tested Cons: implantation technique surgically more demanding than ONS, up to now high rate of lead complications, more invasive.
Taken together, the good or potentially better efficacy of hcSCS has to be weighed up against the higher probability to encounter a lead complication in the course of time. Electrode complications generally are relatively easy to handle; they require a short intervention in local anaesthesia. Therefore we feel that the term “major adverse” event for a lead breakage is misleading, but we admit that this expression is not clearly defined.
However, a patient who under neurostimulation is pain free or nearly pain free will probably be much more affected by a non-functioning neurostimulation device than a patient who has only small benefit from neurostimulation. In our view, one important point is whether hcSCS may also be effective in patients who did not profit from ONS, thus whether stimulation of the central nervous system exerts an additional effect.
We feel strongly that given the increasing importance of less invasive techniques (ONS, hcSCS, eventually stimulation of the sphenopalatine ganglion (6)), hypothalamic stimulation should only be considered as last reserve treatment.
Anyway, future applications of hcSCS, regardless of whether it is used as first line or reserve therapy after unsuccessful ONS, should be performed in the setting of a clinical trial to provide more comparable data to the clinical community.