Trigeminal Autonomic Cephalalgias

Introduction

Trigeminal autonomic cephalalgias (TACs) are a group of headache disorders characterized, as the name suggests, by autonomic symptoms that accompany the headache. These symptoms commonly include rhinorrhea, nasal congestion, conjunctival injection and lacrimation; less common symptoms include eyelid edema, miosis, ptosis, facial erythema and facial perspiration. Trigeminal autonomic cephalalgias comprise four headache disorders:

• Cluster headache (CH)
• Paroxysmal hemicrania (PH)
• Hemicrania continua (HC)
• Short-lasting unilateral neuralgiform headache attacks (SUNHA), with conjunctival injection and tearing (SUNCT) or with cranial autonomic symptoms (SUNA).

All TAC disorders are unilateral, and the most important distinguishing feature is the duration and frequency of the headache attacks, which range from continuous in HC to progressively shorter but more frequent patterns in CH, PH and SUNCT/SUNA (Figure 4-1).

Enlarge 

Cluster Headache

Cluster headache is the most common TAC, but is much less prevalent than migraine, affecting ~0.1% of people in the general population. Historically, CH has been known by many names, including, among others:

• Horton’s headache
• Ciliary or migrainous neuralgia
• Histaminic cephalalgia
• Vidian and Sluder’s neuralgia.

In 1952, Charles Kunkle and colleagues coined the term “cluster headache” to describe the characteristic periodicity of these headaches, which present in series, or so-called “cluster” periods (Figure 4-1).

The ICHD-3 describes CH as attacks of severe, strictly unilateral pain, which is orbital, supraorbital, temporal or in any combination of these sites, lasting 15-180 minutes and occurring from once every other day to eight times a day.

The ICHD-3 diagnostic criteria for CH are shown in Table 4-1. Episodic CH fulfils the general CH criteria and occurs in cluster periods which usually last between two weeks and three months. Chronic CH fulfils the general CH criteria and is characterized by attacks occurring for at least one year either without remission or with remission of less than three months.

Three pathophysiological pathways mediate the clinical presentation of CH. The trigeminovascular pathway (Figure 4-2a) involves cortical structures involved in pain processing, leading to the severe pain associated with CH. The trigeminal-autonomic reflex (Figure 4-2b) involves structures in the pons which are connected to parasympathetic nerves that innervate facial glands, leading to the characteristic lacrimal, nasal and pharyngeal symptoms. Finally, the hypothalamus (Figure 4-2c), which receives nociceptive input from the spinal trigeminal nucleus neurons, may play a role in CH attack generation and is implicated in the circadian and circannual patterns of CH attacks.

The pain of CH is noted for its intense severity. It is described as:

• Throbbing or pulsating
• Severe
• Constant.

Some patients will describe the pain as sharp, knifelike, or stabbing sensations in the eye. The excruciating nature of the pain has prompted some patients to attempt or commit suicide. Patients desire to be upright and moving during an acute attack of CH, as opposed to patients with migraine who typically prefer to be quiet and resting during a severe attack.

The duration of acute CH attacks is noted for its brevity, lasting from 15 minutes up to 3 hours (but typically 30-60 minutes). Patients will experience one to several attacks per day during a series. Many patients will be awakened at the same time each night.

The cluster periods occur most frequently in the spring and fall. During a cluster series, the patient may note that alcohol can trigger an acute attack, although it will have no effect during a remission period.

Cluster headaches, in contrast to migraine, are primarily a male disorder. The prevalence rate in males is believed to be 10:1. The onset of CH usually occurs between the ages of 20 and 40 years. There is an infrequent reporting of a hereditary history of CH. Table 4-2 compares the features of cluster and migraine headaches.

Enlarge 

Enlarge 

Enlarge 

Cluster Headache: Acute Treatment Options

Treatment of CH can be categorized into acute (for ongoing attacks), interim (intended to reduce attack severity and/or frequency before long-term preventive therapy), and preventive (prophylactic); many of the currently available options are presented in Table 4-3. The priority of use for some of these agents is shown in Table 4-4.

The characteristic limited duration of acute CH limits the number of agents used in abortive therapy. In the modern era, these agents include:

• Oxygen
• Sumatriptan
• Zolmitriptan.

For patients whose CH attacks are refractory to pharmacological therapy, two non-pharmacological methods have demonstrated some success in acute CH treatment:

• Vagus nerve stimulation
• Sphenopalatine ganglion (SPG) stimulation.

Oxygen (100%) inhalation at 7-12 L/minute, by facial non-rebreather mask for 10 to 15 minutes, has demonstrated efficacy in abortive therapy of acute CH attacks, and has been used since the 1950s. Multiple studies have reported an attack abortion efficacy of 56-78% within 15 minutes of initiating high-flow oxygen.

Sumatriptan has also shown efficacy for CH abortive therapy. The rapid onset of action with subcutaneous (SC) sumatriptan (6 mg) enhances its benefits in acute CH; in one placebo-controlled trial, SC sumatriptan resulted in severity reduction in 74% of CH attacks, compared to 26% with the placebo. The short half-life (2 hours) of oral sumatriptan precludes its use in acute cluster headache. Intranasal sumatriptan (20 mg), with an initially higher plasma concentration than its oral preparation, is more effective in the management of acute CH.

Intranasal zolmitriptan (5 mg or 10 mg) also has evidential support for abortive treatment of CH attacks in two randomized placebo-controlled trials. Compared to the placebo groups, in which 30% of patients achieved an improvement in headache at 30 minutes, 63% and 48% of patients in the zolmitriptan 10 mg and 5 mg groups, respectively, achieved an improvement in headache.

Two non-pharmacological treatments also have evidential support for efficacy in acute CH. In the Acute Treatment of Cluster Headache 1 (ACT1) study, non-invasive vagus nerve stimulation achieved better pain reduction within 15 minutes (26.7% of attacks) compared to sham stimulation (15.1%). In the ACT2 study, there was no significant difference between the real and sham stimulation in the overall patient population or in the chronic CH subpopulation, but in the episodic CH subgroup, the real stimulation was significantly more efficacious in pain reduction (48% of attacks) than the sham stimulation (6%). In another two clinical trials, SPG stimulation was also superior to sham stimulation with respect to pain reduction in the first 15 minutes (67% vs 7% of attacks) and pain relief (62.5% vs 39% of attacks).

Pain-relief measures may also be considered. However, the acute attack may have subsided by the time the patient reaches a physician’s office or emergency department to receive an analgesic. The use of transnasal butorphanol may offer a rapid, easily administered alternative to parenteral analgesics.

Topical or intranasal capsaicin is available over the counter for pain relief. Although capsaicin does not affect the headache process, it may relieve the acute pain. Intranasal capsaicin appears to be more effective in patients with episodic than chronic CH. In the author’s clinical experience, the results of this therapy have not been promising.

Interim treatments for CH are intended as a bridge treatment before prophylactic therapy can start having an effect. Interim treatments include prednisone and greater occipital nerve block. Corticosteroids, including prednisone, have been in use for the treatment of CH since the 1950s. More recently, a 5-day course of 100 mg daily prednisone in combination with gradually up-titrated verapamil was shown to be superior to up-titrated verapamil alone reducing the number of episodic CH attacks in the PredCH trial. The anti-inflammatory effects of the corticosteroids probably explain their effective use in CH treatment. Corticosteroids should be gradually tapered and eventually discontinued (Table 4-5). The other interim treatment for CH, greater occipital nerve (GON) blockade – the injection of corticosteroid and an anesthetic into the GON – has been used as interim treatment in CH since 1985 and is generally effective, particularly in episodic CH.

Enlarge 

Enlarge 

Enlarge 

Cluster Headache: Prophylactic Therapy

Table 4-3 reviews both abortive and prophylactic agents used in CH therapy. The goal of prophylactic therapy is to limit the length of the cluster period and decrease the severity of the attacks that occur. Agents of choice in the prophylactic therapy of CH include:

• Verapamil
• Lithium carbonate
• Gabapentin
• Galcanezumab.

Non-pharmacological options for prophylaxis of CH include occipital nerve, SPG and vagus nerve stimulation, and CT-guided radiofrequency thermocoagulation of the SPG.

Enlarge 

Verapamil

Calcium channel blockers have been recognized as effective in the treatment of both episodic and chronic cluster headaches. The action of these agents in cluster headaches is not attributed to an effect on vascular dilation. Researchers have implicated the blockade of the release of the pain-inducing neurotransmitters such as substance P, which has been identified in the pathogenesis of CH. Substance P is dependent on calcium for its release. The use of the calcium channel blocker verapamil for chronic CH prophylaxis has good evidential support; a meta-analysis of open-label studies has shown that 87% of patients on verapamil either eliminate attacks entirely or reduce attack frequency by 50% or more.

Lithium Carbonate

Lithium carbonate has been demonstrated by both Ekbom and Kudrow as effective therapy in chronic cluster headaches. Although the method of action of lithium in this disorder is not fully understood, some investigators have attributed this efficacy to alteration of electrical conductivity in the central nervous system. The usual dosage of lithium is 300 mg three times per day. To avoid toxicity, the serum lithium level should be monitored every 4 to 6 weeks and maintained between 0.5 and 1.5 mEq/L. If the patient’s drug levels cannot be easily monitored, another therapeutic agent should be considered. Transient, mild side effects associated with lithium therapy include thirst, polyuria, fatigue and tremor.

Galcanezumab

Galcanezumab is a humanized IgG4 monoclonal antibody directed against the CGRP ligand, which prevents its association with the CGRP receptor. It is one of the newer agents for the treatment of headache, having received initial FDA approval in September 2018 for the preventive treatment of migraine; in June 2019, galcanezumab received FDA approval for the treatment of episodic CH, on the basis of positive results from a phase 3 trial of 106 patients with episodic CH. Patients in the galcanezumab group demonstrated a mean reduction in the number of weekly CH attacks of 8.7, compared to 5.2 in the placebo group (P = 0.04). At week 3, more patients in the galcanezumab group (71%) also achieved a reduction of 50% or more in headache frequency compared to the placebo group (53%; P = 0.046). Galcanezumab is administered by SC injection, typically by the patient. In the phase 3 trial, injection-site pain was more common in the galcanezumab group. Galcanezumab is available as a single-dose pre-filled 120 mg pen or 120 mg or 100 mg single-dose, pre-filled syringe. The recommended dose of galcanezumab is 240 mg initially (as a loading dose), and 120 mg once per month thereafter.

Other Pharmacological Treatment Options

Melatonin – a hormone involved in circadian rhythm synchronization – has been investigated in CH under the rationale that CH attacks often exhibit a circadian periodicity. The only data supportive of melatonin use comes from a small randomized placebo-controlled trial in 20 patients (of which 18 had episodic CH and 2 had chronic CH); melatonin (10 mg daily) demonstrated a reduction in headache frequency, compared to the placebo.

The use of topiramate – an anti-epileptic drug – in CH prevention is supported by case series and two small open-label studies, with reported remission efficacy of 77% in patients on doses of 100-400 mg daily. However, the total number of patients was too small to draw a definitive conclusion, and the use of topiramate remains limited by its adverse events, including cognitive impairment, teratogenicity, nephrolithiasis and mood effects.

Finally, the anticonvulsant gabapentin showed some efficacy for CH prevention in two small trials in patients with CH refractory to other pharmacotherapy.

Non-pharmacological Treatment Options

Several methods of neurostimulation have been investigated for the preventive treatment of CH. In the ICON trial, occipital nerve stimulation of patients with chronic CH at both 100% and 30% intensity significantly decreased the weekly attack frequency compared to the baseline. Studies of SPG stimulation for the preventive treatment of CH have shown a 35-55% reduction in CH attack frequency. Noninvasive vagus nerve stimulation was shown to be effective as an adjunct to standard of care therapy in the PREVA trial, demonstrating significantly greater reduction in the weekly number of attacks (-5.9) compared to standard of care therapy alone (-2.1; P = 0.02). Finally, one small study in 10 patients with chronic CH reported a benefit for CT-guided radiofrequency thermocoagulation of the SPG.

Histamine Desensitization

Histamine desensitization has been used in the treatment of patients with chronic cluster headaches who are refractory to other forms of therapy. The therapy is administered via IV solution combined with histamine phosphate. This treatment is best undertaken in an inpatient setting, with the patient closely monitored for effects of the therapy. Treatment can be concluded in 9 to 11 days.

Although histamine desensitization has not gained in popularity for the treatment of cluster headache, it is a viable and safe alternative. During therapy, if the patient experiences headache, facial flushing, or throbbing, “sensitization,” instead of desensitization, may be occurring. In such cases, the next dose should be reduced by one half, and one should then subsequently build it up again to immediately below the level at which the reaction started.

In a study at the Inpatient Diamond Headache Clinic in Chicago, improvement was noted in 48 of 59 patients. The treatment protocol for IV histamine desensitization at the inpatient unit of the Diamond Headache Clinic is as follows:

• Day 1: 2.75 mg histamine phosphate is used in a slow drip lasting 8 hours
• Days 2 through 11: the patient receives two bottles each day containing histamine phosphate 5.5 mg in each bottle (total of 21 bottles)
  • Start each bottle at the rate of 20 cc/hour
  • Increase by 10 cc/hour every 15 to 30 minutes until patient experiences symptoms of flushing or stuffiness (not to exceed 125 cc/hour)
  • Titrate the drip rate according to the patient’s tolerance. If the patient develops symptoms of headache, decrease the IV rate immediately.

Surgical Intervention

For patients refractory to all types of therapy, surgical intervention may be considered. Some surgical procedures have been abandoned, such as:

• Resection of the greater superficial petrosal nerve
• Section of the nervus intermedius
• Ablation of the sphenopalatine ganglion.

Procedures involving the trigeminal ganglion have been proposed. These procedures include:

• Partial trigeminal root section (posterior approach)
• Percutaneous radiofrequency gangliorhizolysis
• Glycerol injections of the trigeminal cistern
• Posterior fossa trigeminal sensory rhizotomy.

Surgical therapy for CH should only be considered in those patients who have not responded to any conventional form of therapy.

Other Forms of TACs

Beside CH, three other TACs – PH, HC and SUNCT/SUNA occur with comparable (rare) prevalence in the general population. All share common pathophysiological mechanisms in the activation of the trigeminal–autonomic reflex and hypothalamic involvement (Figure 4-2), but are distinguished by attack duration and periodicity (Figure 4-1).

Hemicrania continua is likely the second most common TAC, after CH; unlike CH, it is more common in women. Unlike other TACs, which are characterized by distinct headache attacks, HC is distinguished by continuous mild or moderate baseline pain, with episodes of pain exacerbation. The associated autonomic features are ipsilateral. Hemicrania continua exhibits an absolute response to indomethacin, a characteristic it shares with PH. Hemicrania continua can be divided into two subtypes: the more common unremitting subtype, in which the pain never stops for more than 24 hours, and the remitting subtype, in which the pain is occasionally interrupted by bouts of remission lasting 24 hours or more. The ICDH-3 diagnostic criteria for HC and its subtypes are shown in Table 4-6.

Paroxysmal hemicrania is characterized by pain and autonomic symptoms similar to those of CH. The main difference is that PH attacks are shorter (mean duration 26 minutes, compared to 30-60 minutes on average in CH) but more frequent (5-40 per day, compared to 0.5-8 per day in CH); the autonomic symptoms are also typically less severe than in CH. Paroxysmal hemicrania is very rare, with an estimated prevalence of 0.02% is the general population. Like in HC, the pain associated with PH is ipsilateral to the autonomic symptoms, and attacks of PH respond absolutely to indomethacin. Paroxysmal hemicrania can be divided into two forms: episodic, in which bouts of headache attacks occur with 3 or more months of remission between them, and chronic, in which the attacks are either continuous or separated in bouts with less than 3 months remission between them. The ICHD-3 diagnostic criteria for PH are shown in Table 4-7.

The last major TAC type are short‐lasting unilateral neuralgiform headache attacks. As the name suggests, SUNHA are short, lasting 60 seconds on average. However, they occur with high frequency, with a mean of 28 attacks per day. The attacks may occur individually, in groups, or in a saw-tooth pattern of superimposed attacks. Short‐lasting unilateral neuralgiform headache attacks are further categorized into two major types: SUNCT, characterized by autonomic symptoms of conjunctival injection and lacrimation, and SUNA, in which there is either conjunctival injection or lacrimation (or neither) but not both. Both SUNCT and SUNA are further subdivided into episodic (occurring in bouts lasting from seven days to a year, separated by at least 3 months of remission) and chronic (occurring for more than one year with remission periods lasting less than 3 months). The diagnostic criteria for SUNHA and their subtypes are shown in Table 4-8.

Enlarge 

Enlarge 

Enlarge 

Enlarge 

Enlarge 

Other Forms of TACs: Treatment

Non-CH TACs are either continuous (HC) or characterized by attacks of such short duration (PH and SUNHA), which makes acute therapy generally not feasible; thus, treatment for non-CH TACs is preventive. The preferred treatment options for non-CH TACs are summarized in Table 4-9.

Hemicrania continua responds absolutely within a few weeks to indomethacin. For patients with contraindications or intolerance to indomethacin, other pharmacologic treatment options for HC include: topiramate (alone or in combination with indomethacin, as a sparing agent), celecoxib, piroxicam, melatonin in combination with indomethacin (as a sparing agent), corticosteroids, high-dose ibuprofen, aspirin, amitriptyline, acemetacin, verapamil, gabapentin and onabutulinumtoxinA. Non-pharmacological options include greater occipital or supraorbital nerve blockade, occipital nerve stimulation and vagus nerve stimulation.

Like HC, PH responds absolutely to indomethacin, which must be taken daily to prevent relapse. If a patient cannot tolerate or has contraindications to indomethacin, alternative pharmacological options with some evidential support include COX₂ inhibitors, topiramate, verapamil and carbamazepine. Non-pharmacological alternatives include SPG blockade, occipital nerve stimulation and vagus nerve stimulation.

The approach to the prophylactic treatment of SUNHA can be either short-term or long-term. For short-term prevention (up to 12 weeks), the only pharmacological option with good evidence of efficacy is intravenous lidocaine. For long-term prevention, pharmacological options with evidential support include topiramate and lamotrigine, and to a lesser extent gabapentin, carbamazepine, oxcarbazepine, duloxetine, and zonisamide. Non-pharmacological options with some evidential support include occipital, infraorbital and supraorbital nerve blockade, trigeminal microvascular decompression, stimulation of the greater occipital nerves, and deep brain stimulation.

Enlarge