div class=”ChapterContextInformation”>
19. Future Therapies for Trigeminal Autonomic Cephalalgias: Cluster Headache and Related Conditions
The trigeminal autonomic cephalalgias (TACs) are a group of primary headache disorders linked by usually prominent cranial autonomic features [1] that when present are typically lateralized to the side of the pain [2]. The TACs are grouped under section 3 of the current International Classification of Headache Disorders-3 [3]. They consist of cluster headache [4], paroxysmal hemicrania [5], short-lasting unilateral neuralgiform headache attacks with conjunctival injection and tearing (SUNCT)/cranial autonomic feature (SUNA) [6] and hemicrania continua [7]. These are devastating problems with patients describing the pain of cluster headache as the worst they have ever experienced [8].
The current treatments for cluster headache are less than ideal, be it use limits and vascular issues with triptans [9] or efficacy and tolerability issues with medicines such as verapamil, lithium and topiramate [2]. Few, least of all patients, would argue that new therapies are not required. We review here treatments on the horizon, either just arrived or close by, which provide real optimism that we can manage patients with these disorders much better in the near future. We will address the developments by condition since the treatments for the disorders are one of the more important distinguishing features.
19.1 Cluster Headache
Broadly the treatment of cluster headache (CH) can be considered as either acute, i.e. treating the immediate attack, or preventive, the later short-term bridging or medium- to long-term prevention.
19.1.1 Acute Attack Treatment
All patients with CH require an acute therapy, or at least a discussion of options. Current widely used treatments include triptans; serotonin 5HT1B/1D receptor agonists; sumatriptan 6 mg s/c, sumatriptan 20 mg IN or zolmitriptan 5 mg IN; or inhaled oxygen 100% 12–15 L/min [2]. When considering new approaches, one way to do this is by considering the limitations of the current therapies and how new approaches may help.
19.1.1.1 Can We Make Oxygen Delivery More Efficient?
The currently accepted approach to oxygen therapy in acute cluster headache is described as high flow, 12–15 L/min, and was established evidentially by a randomized placebo-controlled double-blind multi-attack crossover study as effective [10]. One approach that is being explored to improve the performance of oxygen is “ultra-high flow” delivered by a demand valve where inspiratory effort alone limits flow rate. There have been some reported advantages, including patient preference [11], although this method would be well served by a rigorous study.
19.1.1.2 How Do We Treat More Than Two Attacks a Day?
A common problem in practice is patients who have more than two attacks a day. If they have three and respond to zolmitriptan NS, this seems a reasonable solution [12]. However, if they do not, or have more attacks, there can be a problem. Oxygen can be used for any number of attacks although it does not always work and is certainly a cumbersome approach in many ways. Smaller doses per attack of sumatriptan s/c may be used [13]; this area deserves further consideration. One important new addition to acute cluster headache therapy is non-invasive vagal nerve stimulation (nVNS), using the gammaCore device. This delivers a proprietary electrical signal consisting of five 5000-Hz pulses repeated at a rate of 25 Hz. A typical dose is a 120 s of stimulation. There are now two randomized sham-controlled studies that demonstrate superiority at 15 min on the pain-free outcome [14, 15]. The device is well tolerated and there is no limit on daily dosing. Indeed repeated dosing may have some preventive effect [16]. Interestingly, the positive effect on acute attacks was only seen in episodic cluster headache not chronic cluster headache [17].
19.1.1.3 My Patient with Cluster Headache Has Significant Cardiovascular Disease; with What Do I Treat Acute Attacks?
In a clinical cohort of middle-aged, often cigarette smoking males, this is not an uncommon problem. Triptans may be relatively or absolutely contraindicated in such patients because of increased cardiovascular risk [9]. While oxygen is an obvious way forward, again it does not work for everyone and has important logistic limitations. Again for episodic cluster headache, acute attacks may be treated with non-invasive vagal nerve stimulation (nVNS) for which there is clear randomized controlled trial evidence [17]. Based on a study that demonstrated octreotide 100 mcg s/c to be more effective than placebo at 30 min [18], pasireotide [19] is currently being explored for the acute treatment of cluster headache (NCT02619617). Pasireotide has a different receptor binding pattern being high affinity for the somatostatin receptor (SSTR)-5 and less so for 1, 2 and 3, whereas octreotide binds mainly to SSTR-2 [20]. SSTR activation has no known or observed vascular effects so that, if effective, it would be a welcome addition to our options for treating acute cluster headache.
19.1.2 Preventive Treatments for Cluster Headache
Most patients will benefit from at least short-term preventive approaches in cluster headache, and some certainly require long-term prevention. For short-term prevention, typical choices are greater occipital nerve region injection (GONi) with local anaesthetic and a corticosteroid [21, 22] or oral corticosteroids [23]. The former is not universally effective, and the latter has the issue of potential osteonecrotic consequences [24]. Patients with chronic cluster headache need preventives in the longer term. Current approaches with verapamil, lithium, topiramate or melatonin have their many limitations [4].
19.1.2.1 What Can I Use for Short-Term Prevention in Cluster Headache?
In patients who have failed previous GONi or may be unsuitable, there were few realistic choices. Some have advocated short-term nocturnal ergotamine [25] or a more modern version, frovatriptan [26]. These have limitations, including either the issue of concomitant cardiovascular disease or the relative contraindication of concomitant triptans. It has been shown that either spontaneous [27] or nitroglycerin-triggered [28] acute cluster headache attacks are associated with elevated levels of calcitonin gene-related peptide (CGRP). Monoclonal antibodies to CGRP have been tested now extensively in migraine and are effective attack preventives [29]. Galcanezumab, a CGRP monoclonal antibody [30], was tested in a randomized placebo-controlled double-blind study in episodic and chronic cluster headache. It was administered s/c monthly for two doses; at the primary endpoint of weeks 1–3, 75% of galcanezumab patients had a ≥50% reduction in attacks compared to baseline [31]. The treatment was well tolerated with no new adverse events than injection site pain, as reported in controlled trials in migraine [30, 32–35]. Interestingly, there was no significant effect in chronic cluster headache [31]. Similarly, a press release reports that fremanezumab, a CGRP monoclonal antibody effective in migraine prevention [36, 37], in a study in chronic cluster headache (NCT02964338) has been stopped for futility; the episodic cluster headache study (NCT02945046) continues.
19.1.2.2 What Can I Do for Patients with Medically Refractory Chronic Cluster Headache?
There may be no more suffering a patient than those with medically refractory chronic cluster headache [38, 39]. When medicines have failed, clinicians have typically turned to invasive approaches. Radiofrequency lesions [40] or Gamma Knife [41] of the trigeminal ganglion has been used, either important side effects, such as anaesthesia dolorosa, or an outcome no better than natural history, respectively. The sphenopalatine ganglion has been ablated [42, 43] with modest outcomes. The trigeminal nerve root has been sectioned [44] with complications including death. These procedures have been sensibly abandoned. Deep brain stimulation of the region of the brain at the posterior most portion of the hypothalamus, which is active in cluster headache [45], have been reported as being useful [46, 47], although brain surgery has serious, albeit rare, morbidity [48]. Notably a randomized controlled trial was negative [49]. The advent of a less specific, yet safer approach, occipital nerve stimulation [50, 51], made it in turn the preferred option, although lead migration, infection [52] and longer-term battery replacement issues and reduced efficacy have been issues.
The sphenopalatine ganglion, which sits in the pterygopalatine fossa draped across the maxillary branch of the trigeminal nerve [53], is a logical target for the treatment of cluster headache. An important component of the pathophysiology of acute cluster headache attacks is activation of the trigeminal autonomic reflex [54], which accounts for the cranial autonomic features, such as lacrimation, conjunctival injection, nasal congestion, aural fullness and periorbital oedema. The outflow pathway for these symptoms traverses the facial, VIIth, cranial nerve and synapses in the sphenopalatine ganglion (SPG) [55]. Based on this anatomy and clinical experience that the SPG may be a therapeutic target [43, 56, 57], an SPG microstimulator has been developed. In a study, CH-1, comparing SPG stimulation to a sham with no stimulation and a sub-perception stimulus in a randomized crossover design in 32 subjects with chronic cluster headache, 67% of attacks have pain relief at 15 min compared to 7% for each of sham and sub-perception treatments [58]. Interestingly they also reported that 36% had a ≥50% reduction in attack frequency [58]. Most recently the CH-2 study compared SPG stimulation to a sham stimulation that produced a cutaneous TENS-like effect to preserve blinding in subjects with chronic cluster headache. SPG stimulation was more effective than sham at achieving pain relief at 15 min (odds ratio 2.62) and reduced weekly attack frequency by 50% in subjects on active treatment and 28% on sham stimulation [59]. There were no serious adverse events, save surgical events that all resolved. Long-term open-label experience demonstrates that most patients, attack responders or frequency responders, maintain benefit out to at least 24 months [60].
19.2 Indomethacin-Sensitive TACS: Paroxysmal Hemicrania and Hemicrania Continua
Of the TACs, paroxysmal hemicrania (PH) and hemicrania continua (HC) can be very rewarding to treat or remarkably frustrating. Patients respond to indomethacin, by definition. When indomethacin cannot be tolerated, one can use medicines such as topiramate or melatonin, although none are spectacular as indomethacin is.
19.2.1 What Can I Treat Indomethacin-Sensitive Headache Patients with when They Do Not Tolerate Indomethacin?
When patients with PH or HC cannot tolerate indomethacin, their quality of life reverses quickly to that prior to diagnosis and treatment. We have seen of nine patients with HC who could not tolerate indomethacin, seven reported a positive effect on pain with non-invasive vagal nerve stimulation (nVNS) [61]. Similarly, of six patients with PH, four reported significant benefit for nVNS [61]. It is a well-tolerated approach and offers much to patients with very limited treatment options.
19.3 Comments
The diversity in modes of treating TACs and their relatively low success rate point towards the need of rethinking the pathophysiology. The drugs and procedures listed above are by and large focusing on targets outside the CNS. The modulation of the cranial nerves suggests some common pathways. With that in mind future neuroanatomy and function studies will provide avenues for basic research that can help this group of subjects with severe and long-lasting pain syndromes.
