Abstract
The past 5 years has seen therapeutic use of occipital nerve stimulation (ONS) for occipital neuralgia extend to include a range of primary headaches, including chronic migraine (1), cluster (2) and hemicrania continua (3). The linked article by Reed and colleagues reminds us that ONS is a form of peripheral nerve stimulation (PNS) and explains how a ‘paradigm shift’ occurred in 2003 during therapeutic use of ONS for primary headache by moving away from the generally accepted clinical approach of concordant paraesthesia: paraesthesia covering the painful region. ONS produces occipital paraesthesia, but the pain of primary headache is most commonly perceived in the trigeminal region, hence not concordant. The article goes on to describe a study suggesting that a higher responder rate (percentage of patients with a ≥ 50% improvement) could be achieved if the former approach, i.e. concordant paraesthesia, was readopted when using PNS for primary headache.
Reed and colleagues initially explain how the majority of ONS studies have involved three diagnostic categories: ‘occipito-cervical’ with pain in occipital nerve territory, and two primary headaches (migraine and cluster) with fronto-temporal pain in trigeminal nerve territory. Two studies for ‘transformed migraine’ (4,5) were placed in the ‘occipito-cervical’ category as patients had occipitally focused pain within a C1–3 distribution (Reed and colleagues discussed this with Alo, co-author on both papers). The diagnosis of ‘occipital neuralgia’ (6) was also retained for patients, despite reports most were reclassified as chronic migraine (7–9). Using these categories, the average ONS responder rate has been 88% for occipito-cervical but 40–50% for migraine and cluster headache.
Reed and colleagues hypothesized patients with hemicranial/holocephalic primary migraine would respond better to ‘dual’ occipital and supraorbital nerve stimulation (SONS), as opposed to ONS alone. Eight patients with chronic migraine who had failed medical management under the care of experienced headache specialists were recruited. During a 3–5-day trial, all patients received bilateral dual stimulation with two programs: one providing occipital nerve alone and the other dual stimulation. Seven patients (88%) responded with 80–100% improvement in headache while using dual, but no patients responded to ONS alone. Electrodes for bilateral dual stimulation and a pulse generator were implanted in patients who had responded, and the same beneficial responses continued during a median follow-up of 15 months (range 1–35). All patients were able to withdraw or reduce medication and return to fully functioning lifestyles. Adverse events included a single lead migration, an infected lead and an allergic reaction to the pulse generator.
Although successful use of dual stimulation has been reported for neuropathic craniofacial pain (10,11), Reed and colleagues are the first group to report effective dual stimulation for primary headache. However, as conceded by the authors, the study was small, not randomized and open-label with a retrospective assessment of outcomes without diaries. Despite these limitations, the study supports the suggestion of further studies of dual stimulation for primary headache.
One question not asked by Reed and colleagues was whether isolated SONS, without ONS, would be effective for primary headache. SONS has been reported to be beneficial for one patient with chronic cluster headache (CCH) (12), but PNS for primary headache has mainly been studied as ONS for chronic migraine and cluster headache. Several clinical trials of ONS for chronic migraine are in progress (NCT00286078; NCT00747812; NCT00200109), but studies of CCH have been limited to case series (see linked article for references). The answer to why SONS for primary headache has not been studied to the same extent as ONS is not clear. Possible reasons for not studying SONS may have included:
a group of patients with occipital neuralgia who had responded to ONS were later reclassified as chronic migraine, suggesting ONS alone could offer a high responder rate for migraine (see categories of ONS studies, above); previous interchangeable use of occipital neuralgia and ‘transformed migraine’ (13,14) may have influenced opinion as to the potential benefits of ONS for primary headache; in the first half of the last decade the vast majority of safety data were for ONS and there was a lack of data for SONS, although an assumed higher risk of infection or skin breakdown for SONS may not be real (15).
Despite ongoing investment in the study of ONS for primary headache, if further studies of dual stimulation for migraine also achieve high responder rates, larger controlled studies would then seem likely, despite inherent blinding issues for PNS due to patients' perception of paraesthesia. The study of dual stimulation for CCH should also be considered, because high responder rates (compared with ONS) may prevent some patients having to consider deep brain stimulation (DBS) (16,17), which would be a definite advantage as DBS carries a risk of serious adverse events (18). However, DBS may still be offered to some patients who fail to respond to PNS, or if DBS continues to offer a greater chance to become pain free/almost pain free.
A question arising from the linked article is what mechanistic advantage might dual stimulation have over ONS in terms of modulating the pain of primary headache, in order to produce higher responder rates? Reed and colleagues speculate that partial convergence of ipsilateral trigeminal and cervical neurons in the trigeminocervical complex (19) may play a part. In attempting to understand the mechanism of action of ONS for primary headache, we start without knowing the origin of pain in primary headache, which is considered to occur despite ‘neither bodily nor neuronal injury’ and therefore not classified as nociceptive, neuropathic or inflammatory (20,21). One approach to understanding the mechanism of action of ONS for primary headache has focused on identifying its site of action, and studies have suggested central segmental or higher level processing may play a role (22,23). Another study used a neuropharmacological approach to look for opioid or GABA mechanisms for ONS efficacy, but no clear evidence for either was obtained (24). There is more work to do in trying to understand the mechanisms responsible for the improvement in headache during use of ONS, although dual stimulation may provide us with insight to further our understanding of PNS mechanisms and possibly the pathophysiology of primary headache.
Although the linked article is a small open-label study, the results suggest use of dual stimulation may offer patients with refractory migraine a therapeutic benefit over ONS. Provided diagnostic accuracy is maintained and if the high responder rate is consistently reproduced in larger dual stimulation studies (ideally placebo controlled), without excessive adverse events, this would change our current approach to PNS for primary headache.