A review of the current and potential oxygen delivery systems and techniques utilized in cluster headache attacks

Abstract

Background

Despite oxygen therapy being one of the foremost acute treatments for cluster headache (CH) attacks, little is known about the different techniques and systems.

Objectives

In this review we will examine the efficacy of the standard non-rebreather mask (NRM) with room temperature oxygen in relieving pain in CH, and try to compare it with the diversity of other oxygen gas conditions and interfaces like partial rebreathers, simple masks, nasal cannulas, tusk masks, demand valve oxygen, hyperbaric and cooled oxygen.

Method

We searched non-structured Pubmed, Medline, the Cochrane online database and instruction protocols from various oxygen delivery devices.

Conclusions and implications

Interfaces like demand valves and tusk masks are already proving to be superior or at least similar to the standard NRM in terms of fraction of inspired oxygen (FiO₂), though the demand valve only showed better results than the NRM in a single study in only four participants. Furthermore, new research shows how lower temperatures of the gas may be an essential part of effective pain relief and hyperbaric treatments show potential in preventing night time attacks.

Keywords

cluster headache oxygen delivery system oxygen therapy non-rebreather mask nasal cannula tusk mask demand valve oxygen simple mask hyperbaric oxygen therapy fraction of inspired oxygen CPAP oxygen temperature

Introduction

Cluster headache (CH) is a severe type of headache that can have a debilitating impact on the quality of life for the patient. It has been described as a painful sensation with an intensity worse than giving birth and can incapacitate the patient (1). A number of acute and prophylactic treatments are available for a patient to alleviate the pain, the most common of which being the serotonin receptor agonist sumatriptan during a CH attack and the calcium antagonist verapamil as a prophylactic. While proven to be effective for relieving pain, sumatriptan does have a few drawbacks. Sumatriptan (2) has been known to cause vasospasm of coronary arteries (3). Therefore, sumatriptan has been contraindicated as acute medication for patients with ischaemic heart disease, a history of myocardial infarction or uncontrolled hypertension (4).

Another first line acute treatment is oxygen. While certainly not new, as it has been recommended in CH since 1952 (5), it has steadily increased in use during the past decade. By 2010, in the United States alone, two-thirds of CH sufferers had tried oxygen therapy (6). Approximately 44–61% of CH patients experience relief from oxygen during attacks and several patient characteristics that play a role in its efficacy have been identified (7,8). The proposed underlying mechanisms at play are (the resulting) cerebral vasoconstriction, a modulatory effect on the parasympathetic outflow system, and possibly the anti-inflammatory effects of hyperoxia (9,10). Oxygen flow rates of 6, 7 and 12 L/min have been proven to be effective (11 –13). Oxygen flow rates of 14–15 L/min have also achieved relief in three patients who were resistant to 7–10 L/min (14). Relatively little attention, however, has been paid to technical aspects of oxygen treatment.

As of now, oxygen treatment for CH is applied mostly by non-rebreather masks (NRMs) as the preferred oxygen delivery system (6,15). One of the reasons to favor NRMs over simple masks is because a NRM can deliver higher concentrations of oxygen to the patient with less entrainment or mixing of normal air per inspiration. Applying this technique, it can be assumed that the efficacy of oxygen therapy in CH relies on the fraction of inspired oxygen (FiO₂). FiO₂ itself has a large impact on partial oxygen pressure within the respiratory system and can be used to calculate the eventual alveoli oxygen pressure (P_AO_2; Figure 1) (16).

Figure 1.

PB = Barometric Pressure in mmHg (Sum of partial pressure of gasses in alveolar air) PH₂O = Partial pressure in mmHg of water vapor in alveolar air (Healthy subject = 47) P_aCO₂ = Partial pressure of arterial CO₂ R = Respiratory exchange ratio (Healthy subject = 0.8, this ratio is omitted when FiO₂ is above 60%) (16).

In this review we will look at different oxygen delivery systems, techniques, and conditions. Moreover, we will see how they compare to today’s standard method with respect to rate of FiO₂, comfort, ease of use, adverse effects, and possible pain relieving efficacy on CH attacks.

Methods

We searched non-structured Pubmed, Medline, the Cochrane online database and instruction protocols from various oxygen delivery devices to complete our investigation.

Oxygen masks

Non-rebreather mask

Most clinicians today prescribe a NRM as the standard oxygen delivery system for CH (Figure 2(a)). It can achieve a FiO₂ of 45 to 95% depending on the flow rate used (17 –19). It works by storing oxygen in a reservoir bag of about 1 L (20). As the patient inhales the air from the reservoir bag, this passes through a one-way valve, thereby preventing exhaled air from entering the bag. The clinician should also point out to the patient that the reservoir should be inflated for at least 1/3 of its maximum volume, if it goes below this threshold the flow rate should be increased (21). The NRM always has an oronasal interface. The exhaled air escapes through expiratory portholes on the sides of the mask and most NRMs also include flaps covering the portholes that partially prevent normal air from entering the mask, thereby preventing entrainment of ambient air and lowering the FiO₂ (20). Omission of these flaps is sometimes applied as a precautionary measure in the event that the inspiratory one-way valve malfunctions or the flaps adhere shut by humid air, which would interrupt the flow of oxygen to the patient (17,22).

Figure 2.

(a) Non-rebreather mask, (b) Simple mask, (c) Nasal cannula, (d) High flow nasal cannula (e) Partial rebreather mask, (f) Tusk mask, (g) Demand valve Oxygen, (h) HBOT, (i) CPAP.

Applying the mask is simple and very few patients experience discomfort from the interface itself. Those who do have an apprehension of using a NRM, usually suffer from claustrophobia when using a mask or want their mouth free from obstructions (23).

NRM oxygen flow rates are limited to a range of 6–15 L/min. The current therapeutic oxygen flow rates in CH, being 7–12 L/min (6,12,13), can be delivered from 10 L gas cylinders at 200 to 300 bar or the smaller portable ones of about 2 L (6). The optimal flow for relieving CH pain is still under investigation, however, higher flow rates seem to be more effective in relieving pain (14). In terms of longevity, a typical canister of 10 L can last for about 3.5 hours, and at an average flow rate of 10 L/min, is enough for roughly 14 sessions. NRM has proven effective in relieving pain during CH attacks in several publications since 1985 (11,12).

Normal (simple) mask

Simple face masks (Figure 2(b)) are different from the more commonly used NRMs. They do not allow a high or accurately controlled FiO₂, which is at maximum around 50% at a flow rate of 5–12 L/min (17,19,24). The masks have open expiratory portholes on either side without the flaps that are present in NRMs, unfortunately increasing the entrainment of ambient air, but also negating the complications inherent to the flaps (25).

A simple face mask is usually indicated for short-term use with low discomfort level, in cases of mild hypoxemia. Whether or not actual research has been performed using a simple face mask is unclear, since most publications on normobaric oxygen therapy either state that they used NRMs or undefined facial masks (26). In a survey in the United States, 29% of 1134 participants with CH stated they used a simple face mask (6).

Nasal cannula

Nasal cannulas (Figure 2(c)) are comparable to simple facial masks as they have a simple design and do not allow for accurate control of the FiO₂ (17). The FiO₂ can vary from 22% to 45% depending on the flow rate and minute ventilation of the patient. The flow rates are usually kept low between 0.25 and 6 L/min (17,19). Flow rates higher than 6 L/min are discouraged, because of drying of the nasal mucosa and in more severe cases septal perforation (17). Clinicians with experience in the usage of oxygen therapy in CH often prefer using facial masks or NRMs over nasal cannulas, because of the difficulty to equip nasal prongs during a CH attack. Moreover, the lower flow rate and FiO₂ puts its efficacy in pain relief into question.

In 2010, of 1134 participants with CH in a survey in the United States, about 11% were using oxygen therapy by a nasal cannula (6).

In a different survey performed in Belgium, 69% of the participants with CH was prescribed oxygen therapy. Thirteen percent was administered by way of nasal cannula or used this method in the past. The authors themselves claimed that this prescription was incorrect and preferred the use of a NRM (15).

There is, however, a high flow variant of the nasal cannula (Figure 2(d)) that can operate at flow rates ranging between 15 and 40 L/min with a constant FiO₂ above 80% regardless of the patient’s ventilation rate, thanks to artificially heated and humidified oxygen (17).

Like low flow nasal cannulas, using high flow nasal cannulas for relieving pain in CH attacks has not yet been studied. According to product specialists, high flow nasal oxygen delivery systems are impractical and not cost effective for use in a home setting, where most of the CH attacks occur, since the necessary equipment is costly and difficult to relocate to a patient’s home. Furthermore, the high flow rate would consume oxygen cylinders at a much higher rate than NRMs or simple face masks, adding more to the costs and logistics in providing a sufficient supply of gas cylinders.

Partial rebreather mask

The key difference between the NRM and the partial rebreather mask (PRM; Figure 2(e)) is that the reservoir bag allows the first third of the exhaled air in. This derives mainly from the anatomic dead space that has a higher O₂ concentration than the remaining two thirds that leaves through the ventilation portholes on the mask. This mechanism ensures a higher FiO₂ than the simple (face) mask (17).

The PRM is able to deliver a FiO₂ of 35 to 60% at flow rates of 6 to 10 L/min. The PRM has mostly been replaced by the NRM, since a NRM allows higher levels of FiO₂, (17,19) despite the fact that the PRM experiences less airflow interruptions and resistance due to the partially open inspiratory valve (27).

There is no mention in current literature of the PRM being used to treat CH.

Tusk mask variants

This relatively new method (Figure 2(f)) has been developed to resolve the shortcomings of the NRM, when it comes to providing a consistent FiO₂ and preventing potential expiratory valve problems. In modifying a PRM by adding an open respiratory tube of about 15 cm on each expiratory porthole, an expanded artificial dead space is added to prevent entrainment of ambient air (28). For a tusk mask to function properly, it does require a high flow of oxygen to flush out the exhaled air from the “tusks” and prevent ambient air entrainment. The specific flow rate to accomplish this is unclear (17).

Lambrigts et al. (29) compared the FiO₂ from the PRM and the tusk mask on four different flow rates in 10 healthy subjects. In the trial the tusk mask had a consistently higher FiO₂ when the flow was between 5 and 15 L/min, being almost 50% higher than the FiO₂ values of PRM (see Table 1).

Table 1.

Various oxygen supply methods and characteristics.

Oxygen delivery system	Flow rate (L/min)	FiO₂ (%)	Verified efficacy on CH	(Dis)advantages
NRM (17,19,21,30,33)	6–9	45–55	++	+Multiple studies proving efficacy in CH +Simple to use with little discomfort −Risk of valve obstruction −Requires a high flow rate to keep reservoir bag inflated
10–15	55–95
Normal (simple) mask (17,24,33)	5–12	35–50		+Simple to use during CH attack with little discomfort +Less risk of valve obstruction −Lack of trials proving its efficacy in CH −FiO₂ difficult to predict
Nasal cannula (17,19,33)	0.25–6	22–45		−Lack of trials proving its efficacy −FiO₂ difficult to predict −Cumbersome to apply during CH attack −Low FiO₂
High flow nasal cannula (17,19,33)	15–40	≥80		−Risk of drying nasal mucosa after prolonged use −Costs −Logistically impractical
PRM (17,19,33)	6–10	35–60		+Simple to use with little discomfort −Lack of trials proving its efficacy
Tuskmask (18,29)	5	49		−Lack of trials proving its efficacy in CH
	10	75		−Lack of trials proving its efficacy in CH
	15	91.2
			−Uncertainty about exhalation tube safety −Non-esthetical appearance
DVO (18,30)	5–160	100^a	+	+High FiO₂ −Lack of trials proving its efficacy in CH −Costs −Not widely available −Requires a tight mask seal

Unconfirmed.

FiO₂ per Oxygen delivery system.

References: (17,18,19,21,24,29,30,33).

DVO: demand valve oxygen; NRM: non-rebreather mask; PRM: partial rebreather mask.

The safety of the exhalation tubes compared to the common open portholes and flaps is still uncertain.

Demand valve oxygen

The demand valve oxygen system (DVO; Figure 2(g)) is a new delivery system. It is usually applied in emergency rooms for resuscitation in breathing or non-breathing distressed patients and can deliver 100% oxygen from 5 to 160 L/min (18,30). The valve works, as its name suggests, by only opening when there is a negative pressure present in the mask or the tube, which occurs when the patient is inhaling, in other words when there is a demand for oxygen. The valve closes immediately after inhaling. An interesting feature of the system is that the device can support hyperventilation. The rate of oxygen using a DVO is thereby controlled by the respiratory rate and tidal volume. Therefore the DVO requires a tight mask seal to prevent aforementioned negative pressure from increasing (18). The DVO is available with an oronasal mask or a mouthpiece inhalator. The mask does not have exhalation portholes or flaps as the demand valve itself has an outlet for exhalation (31).

Few studies have tested the performance of the DVO and only one to show the efficacy in relieving pain in CH (30,32). In an open-label crossover design pilot study (2013), Rozen and Fishman (30) showed in all subjects (n = 4) that DVO was able to interrupt a CH attack within 30 minutes and reduce cranial autonomic symptoms. Three of these four patients also received oxygen using a NRM for a subsequent CH attack, however, only two of these subjects were pain free after using the NRM. However, given the small population size, one should not take this as a more valid option above the standard NRM.

Another point to mention is that DVO, while reusable, is priced much higher than other oxygen masks, costing around 500 dollars on the open market, an aspect the prescribing specialist should keep in mind (30).

Hyperbaric oxygen

A more intensive therapy is the use of oxygen in a hyperbaric environment. At a pressure of 300 kPa the small quantity of plasma oxygen in the blood can be increased by up to 300% and diffusion through tissues may also be improved resulting in an arterial and tissue oxygen tension of approximately 2000 and 500 mmHg respectively (34,35). Oxygen bound by hemoglobin, however, shows little to no increase since it is usually already saturated in normobaric atmospheres (34).

Hyperbaric oxygen therapy (HBOT; Figure 2(h)) is a costly endeavor. It is also impractical for use as a rescue treatment for a CH attack, since a patient can only receive HBOT treatment in a hospital or private clinical setting (36,37). Despite appearing cumbersome and impractical in serving as a pain-relieving therapy, reports have shown it to be effective in relieving patients from CH attacks since 1989 (38). Porta et al. (39) and Di Sabato et al. (40) showed that HBOT was able to interrupt a CH attack in 14 of 14 CH patients and six of seven episodic CH patients, respectively. Three of the six responders in the Di Sabato study were also free from CH attacks for four to six days and another three patients had no CH attacks during the follow-up period of two months (40).

There have been other studies investigating the use of HBOT as a preventive treatment for CH (41,42). Nilsson Remahl et al. (41) conducted a study on the efficacy of HBOT compared to hyperbaric normoxic placebo treatment. Sixteen patients were treated with 100% oxygen and/or a placebo normoxic treatment in a hyperbaric chamber for 70 minutes with a 24-hour interval between sessions. They did not show a significant difference between placebo and HBOT in reducing CH attack frequency and intensity (41). Pascual et al. (42) studied the frequency and duration of CH attacks during HBO treatment (10 sessions administered in a hyperbaric chamber for 70 minutes per session), compared to the last (minimum) two weeks before treatment start in four chronic CH patients. One patient did not have any CH attacks until 31 days after his eight day treatment. However, one other patient did not experience any effect (42).

With HBO being too cumbersome and expensive to warrant its use as a valid rescue treatment for CH attacks and proving to be no more effective as a prophylactic than normoxic therapy, we do not recommend it for treating CH.

Continuous positive air pressure

Studies indicate that CH is more prevalent in patients with obstructive sleep apnea syndrome (OSAS) than those without (43 –45). Bouts of hypoxia during sleep are postulated to trigger night-time CH attacks (43). However, it should be noted that a recent study failed to find an association between sleep apnea and CH (46). It has also been hypothesized that dysregulation of the hypothalamic-pituitary-adrenal axis may cause sleep apnea and CH as parallel symptoms (44).

Often used for treating OSAS, continuous positive airway pressure (CPAP; Figure 2(i)) may deliver a hyperbaric pressurized air mixture at an airflow typically ranging from 20 to 60 L/min to the patient and expand the lung volume at the end of expiration. This in turn increases the functional residual capacity. The air pressure ranges from the initial 4 cm H₂O to a maximum of around 20 cm H₂O. CPAP can be given by way of a full face mask, nasal pillow or nasal mask, though the choice of interface usually relies solely on the patient’s preferences. Cases of claustrophobia, nasal congestion and abundant facial hair can impact the choice (23).

CPAP has been proven safe to be used throughout the night and could perhaps be used to prevent CH attacks, particularly nocturnal ones, by way of preventing hypoxia due to OSAS. In a few case reports it has been shown to reduce the frequency of night-time CH attacks. Zallek and Chervin (45) described a case of a 60-year-old patient, who was treated with CPAP for his OSAS, which also immediately reduced his night-time CH attacks, although it did not alleviate his day-time CH attacks (45).

Although just a single case, this may imply that CPAP could be a potential preventive measure against nocturnal CH attacks in patients with OSAS. A downside, unfortunately, is that, although CPAP can be outfitted to deliver close to 100% oxygen, this is discouraged to avoid oxygen toxicity (see below) if used throughout the night (47). As such its FiO₂ is usually 21–24% since it provides an air mixture.

We recommend clinicians to pay attention to a possible history of OSAS in patients with CH, especially nocturnal. Starting a treatment with CPAP could then be considered, but this should be further studied.

Technique

Posture

While there has been extensive research on the effects of posture on pulmonary function, little research has been dedicated to posture and its effects on the efficacy of oxygen treatment in CH. In two publications patients are advised to sit upright, when using oxygen (48,49). Oxygen suppliers themselves also instruct their clients to assume an upright sitting position, when using oxygen.

In a cross-sectional study of 20 men, Hojat and Mahdi (50) showed that the forced vital capacity (FVC), forced expiratory volume in one second (FEV1) and peak expiratory flow (PEF) are at their highest value when sitting upright or standing up.

It should be noted that advising a patient to maintain a certain posture may be fruitless, since CH patients have great difficulty staying still.

Breathing technique

Aside from the position in which to breath, the rate and depth of respiration could also have an effect on the FiO₂ of the mask. Clinicians agree that when using a delivery device with 100% oxygen, depth and rate have little to no effect on the FiO₂ of the patient, however, this only accounts for tidal volumes that are below the supplied flow rate (51). Research shows that voluntary hyperventilation may result in a significant increase of arterial PO₂ level (52).

Another consequence of hyperventilation that could have a positive effect on a CH attack is the resulting hypocapnia. It is known that hypocapnia causes vasoconstriction of the intracerebral arteries (53) and this could have a relieving effect on the CH pain. Hannerz and Jogestrand (54) showed in six of eight CH patients that hypercapnia, by inhaling 6% CO₂ for six minutes, could provoke unilateral mild to moderate pain, though no classical CH symptoms were present, in contrast to the control group who showed no symptoms. Unfortunately, little research has been conducted on the pain-relieving effects of hypocapnia in CH. Mention is made of voluntary hyperventilation being used in the DVO study by Rozen and Fishman (30). In the study, all participants using DVO were asked to hyperventilate at the start of a session; however, this does not exclude a normal respiration rate being used. The risk of syncope during voluntary hyperventilation is present in case of pre-existing cardiovascular disease or defect cerebral blood flow, therefore physicians should keep this in mind when advising patients of this particular breathing technique (55).

While technique may not be a truly significant contributing factor to the efficacy of oxygen treatment, we still recommend that patients should always be instructed on the most optimal way of using their oxygen delivery device, namely by sitting upright. The question of whether voluntary hyperventilation is superior is currently unknown and would require further investigation before clinicians recommend this to their patients.

Oxygen conditions

Temperature

The temperature of the supplied oxygen might influence the severity of a CH attack by causing vasoconstriction via sympathetic nerves and could be a critical component for the efficacy of the oxygen therapy (56,57). A study conducted by McLeod et al. (58) explored this question. Applying a NRM, they administered room air, cooled to approximately 5–8℃, to patients during an attack and compared the resulting level of pain relief with six of the same patients, who were followed up with the standard uncooled oxygen therapy. Surprisingly, the cooled air treatment not only provided similar pain relief in the patients’ CH attacks (85% vs. 83%), but the response rate was also slightly faster than that of the standard oxygen therapy (6–15 minutes vs. 7–15 minutes). This raises the question of the underestimated importance of air temperature and perhaps the overestimated importance of the oxygen concentration (58).

Many neurologists who treat CH patients will recognize patients' experiences with respect to the effect of cold air. However, a good response to a cold environment was found in only 6.7–12.5% in an epidemiological study (7).

Common adverse effects

There are relatively few adverse effects, when using 100% normobaric oxygen therapy compared to pharmacological treatments for CH. Using non-humidified oxygen, dehydration of oral and nasal mucosa can occur, which thins the mucosal wall and makes it more susceptible to inflammation and infection (59). When a patient has prolonged exposure, i.e. longer than 48 hours, to a FiO₂ higher than 50%, he or she does run the risk of oxygen toxicity. This type of toxicity is characterized primarily by pulmonary complaints like retrosternal pain, coughing and dyspnea. These symptoms are due to diffuse damage to both alveolar epithelium and pulmonary capillary endothelium followed by interstitial and intra-alveolar edema, impaired gas exchange, and inflammation of the lungs (47,60,61). A less prevalent side effect of oxygen therapy is retinopathy, which is characterized by retinal thinning (62,63). Such prolonged exposure, however, is unlikely to occur in CH patients.

The use of an inflammable gas in the home setting also presents its own dangers. Per protocol every patient prescribed oxygen is reminded not to smoke or use heat or spark producing sources, when using an oxygen delivery device (64).

Another adverse effect, still undergoing investigation, is a rebound effect occurring after an oxygen treatment session, where the pain returns sooner than expected and/or attacks occur more frequently when using oxygen as acute treatment (65). The literature is, however, scarce here.

Unlike normobaric oxygen therapy, hyperbaric oxygen therapy has unique adverse effects. Barotraumatic lesions due to compression or expansion of enclosed gas volumes may occur in the paranasal sinuses, inner ears, and lungs. Ocular damage is also known to occur in long-term use of hyperbaric therapies. Exposure to 100% oxygen at hyperbaric pressures may lead to hyperoxia and oxygen toxicity within a few hours, compared to the days it takes at normobaric pressures. This is further compounded by the fact that hyperbaric oxygen toxicity debuts with the central nervous system adverse effects collectively called the Paul Bert effect (named after the 19th-century Paris physiologist). Here the cardinal symptoms are generalized seizures (47,60).

Discussion

The aforementioned survey performed by Rozen and Fishman (6) noted that only 28% of prescribers of oxygen therapy specified a specific type of delivery system, while the rest left this up to the patient or the oxygen supplier. While this survey only pertains to US medical specialists, it does show a lack of expertise and knowledge, when it comes to oxygen interfaces for their patients and could also be true for clinicians in other countries. This lack of knowledge could very well be existent, because there has never been a comprehensive investigation in comparing the various oxygen delivery devices. While CH is an uncommon headache type, its severity does warrant a better understanding of how to provide a more effective and faster way of relieving such pain. A large crossover trial between the various interfaces and gas conditions could help in choosing the right delivery system.

Studies have shown that decreased oxygen saturation might actually trigger CH attacks (43,66). So while no definitive causal relation has yet been found, one could postulate that the higher the FiO₂, the higher the amount of oxygenation of the blood, the more effective oxygen therapy can be.

As mentioned above, the delivery system most commonly used is the NRM, due to its high FiO₂ output, ease of use and low costs. Research into efficacy of DVO and tusk masks for CH treatment could be interesting as it may provide an effective replacement, given their high FiO₂ characteristics. This might give us a clearer picture on the role of FiO₂ for relieving CH symptoms. As previously mentioned, DVO was already compared to NRM in a crossover study, however, the number of participants was too small to draw any significant conclusions (30).

It is advisable to investigate the effects and the underlying physiology of cooled oxygen further by comparing it with room temperature oxygen and using larger populations suffering from CH (58).

Breathing techniques, notably sitting or standing upright, can be advised by clinicians to their patients. The benefits and safety risks, such as syncope, of voluntary hyperventilation, however, would require more investigation (55).

Another idea worth further investigation is the potential of CPAP to be taken as a preventive measure against attacks as applied in the case report by Zallek and Chervin (45). Specifically night-time bouts of CH, which make up approximately 75% of all the attacks, could be treated this way (67). Other studies have already demonstrated that CPAP is effective in preventing other types of night-time headaches (68). For effective day-time prevention, research could be done to see if less obtrusive devices like nasal cannulas could be used throughout the day to prevent attacks.

Table 2 provides a summary of our suggestions for future research to improve upon the technique and conditions in oxygen therapy in CH.

Table 2.

Suggestions for future studies.

• A direct comparison, preferably in a prospective randomized controlled cross-over trial, to assess the efficacy on pain relief and comfort of use of the various oxygen therapy interfaces, including: NRM, PRM, Tusk masks, DVO, nasal cannulas.

• More intensive research on the efficacy of cooled oxygen in CH patients in a larger sample, comparing its effects to cooled and uncooled air and oxygen.

• The preventive potential of hyperbaric therapies like CPAP against the onset of nocturnal CH attacks.

• The prophylactic potential of long term oxygen therapy using nasal cannulas against the onset of daytime CH attacks.

• An investigation in the efficacy of voluntary hyperventilation on CH severity when using oxygen delivery devices.

Clinical implications

Little is known about the optimal way of administering oxygen for relieving pain in cluster headaches (CH).

Non-rebreather and demand valve oxygen masks show the most promise in effectively reducing pain in CH, with the tusk mask is still needed to be validated in this regard.

Cooled oxygen might have a higher efficacy in combating cluster headache pain.

Continuous positive air pressure (CPAP) could have a preventive potential against the onset of nocturnal cluster headache attacks.

Footnotes

Acknowledgements

The authors would like to thank E. Winselaar-Frik (Westfalen Medical BV, Deventer, The Netherlands) and R. van der Horst, Pneumonologist/ Intensivist (Atrium MC Parkstad, Heerlen, The Netherlands).

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

References

May

Bahra

Büchel

. Hypothalamic activation in cluster headache attacks. Lancet 1998; 352: 275–278.

Law

Derry

Moore

. Triptans for acute cluster headache. Cochrane Database Syst Rev 2010; 4: CD008042–CD008042.

Tepper

Millson

. Safety profile of the triptans. Expert Opin Drug Saf 2003; 2: 123–132.

Farmacotherapeutisch Kompas. Sumatriptan. Zorginstituut Nederland 2014, https://www.farmacotherapeutischkompas.nl/Preparaatteksten/S/sumatriptan.asp (2014, accessed October 2014).

Haane

Dirkx

Koehler

. The history of oxygen inhalation as a treatment for cluster headache. Cephalalgia 2012; 32: 932–939.

Rozen

Fishman

. Inhaled oxygen and cluster headache sufferers in the United States: use, efficacy and economics: results from the United States Cluster Headache Survey. Headache 2011; 51: 191–200.

Haane

De Ceuster

Geerlings

. Cluster headache and oxygen: is it possible to predict which patients will be relieved? A prospective cross-sectional correlation study. J Neurol 2013; 260: 2596–2605.

Backx

Haane

De Ceuster

. Cluster headache and oxygen: is it possible to predict which patients will be relieved? A retrospective cross-sectional correlation study. J Neurol 2010; 257: 1533–1542.

Akerman

Holland

Lasalandra

. Oxygen inhibits neuronal activation in the trigeminocervical complex after stimulation of trigeminal autonomic reflex, but not during direct dural activation of trigeminal afferents. Headache 2009; 49: 1131–1143.

10.

Schuh-Hofer

Siekmann

Offenhauser

. Effect of hyperoxia on neurogenic plasma protein extravasation in the rat dura mater. Headache 2006; 46: 1545–1551.

11.

Fogan

. Treatment of cluster headache. A double-blind comparison of oxygen v air inhalation. Arch Neurol 1985; 42: 362–363.

12.

Cohen

Burns

Goadsby

. High-flow oxygen for treatment of cluster headache: a randomized trial. JAMA 2009; 302: 2451–2457.

13.

Kudrow

. Response of cluster headache attacks to oxygen inhalation. Headache 1981; 21: 1–4.

14.

Rozen

. High oxygen flow rates for cluster headache. Neurology 2004; 63: 593–593.

15.

Van Alboom

Louis

Van Zandijcke

. Diagnostic and therapeutic trajectory of cluster headache patients in Flanders. Acta Neurol Belg 2009; 109: 10–17.

16.

Chang

. Respiratory Care Calculations, 3rd edn. New York: Delmar Publishers, 1999, pp. 14–14.

17.

Criner

Barnette

D'Alonzo

. Critical Care Study Guide, 2nd edn. New York: Springer, 2010, pp. 27–37.

18.

Bouak, F. Oxygen Therapy: Description and Advances in Oxygen Delivery Systems. Report for Defence R&D Canada, Toronto. Report no. DRDC-TORONTO-TM-2004-112, 17 December 2004.

19.

Green

Frain

Oxygen and Medical Gas Therapy. In: Wyka

Mathews

Rutkowski

(eds). Foundations of Respiratory Care, 2nd edn. New York: Delmar, 2012, pp. 557–597.

20.

Marino

. The ICU Book, 3rd edn. Philadelphia: Lippincott Williams & Wilkins, 2006, pp. 406–410.

21.

Sills

Oxygen and Medical Gas Therapy. In: Sils

(ed). The Comprehensive Respiratory Therapist Exam Review, 6th edn. St. Louis: Elsevier, 2015, pp. 208–208.

22.

Ahrens

Rutherford

Rutherford Basham

. Essentials of Oxygenation: Implication for Clinical Practice, 1st edn. Boston: Jones & Bartlett Learning, 1993, pp. 133–141.

23.

Carron

Freo

Bahammam

. Complications of non-invasive ventilation techniques: a comprehensive qualitative review of randomized trials. Br J Anaesth 2013; 110: 896–914.

24.

Kallstrom

. American Association for Respiratory Care (AARC) Clinical Practice Guideline: oxygen therapy for adults in the acute care facility. Respir Care 2002; 47: 717–720.

25.

Hnatiuk

Moores

Thompson

. Delivery of high concentrations of inspired oxygen via Tusk mask. Crit Care Med 1998; 26: 1032–1035.

26.

Petersen

Barloese

Jensen

. Oxygen treatment of cluster headache: A review. Cephalalgia 2014; 34: 1079–1087.

27.

Farias

Rudski

Zidulka

. Delivery of high inspired oxygen by face mask. J Crit Care 1991; 6: 119–124.

28.

Peruzzi

Smith

. Oxygen delivery: tusks versus flow. Crit Care Med 1998; 26: 986–986.

29.

Lambrigts

Van Trimpont

Duprez

. The tusk mask II, a new way of oxygen administration allowing a higher FiO₂: an experimental model. Eur J Anaesthesiol 2001; 18: 55–55.

30.

Rozen

Fishman

. Demand valve oxygen: A promising new oxygen delivery system for the acute treatment of cluster headache. Pain Med 2013; 14: 455–459.

31.

ISO 13485: 2003. Demand/Resuscitator Valve.

32.

Jay

Tetz

Hartigan

. Portable hyperbaric oxygen therapy in the emergency department with the modified Gamow bag. Ann Emerg Med 1995; 26: 707–711.

33.

American Association for Respiratory Care. AARC clinical practice guideline, In vitro pH and blood gas analysis and hemoximetry. Respir Care 1993; 38: 505–510.

34.

Gill

Bell

CAN

. Hyperbaric oxygen: its uses, mechanisms of action and outcomes. QJM 2004; 97: 385–395.

35.

Bateman

Leach

. ABC of Oxygen: Acute oxygen therapy. BMJ 1998; 317: 798–801.

36.

Tibbles

Edelsberg

. Hyperbaric-Oxygen Therapy. N Engl J Med 1996; 334: 1642–1648.

37.

Abidia

Laden

Kuhan

. The role of hyperbaric oxygen therapy in ischaemic diabetic lower extremity ulcers: a double-blind randomised-controlled trial. Eur J Vasc Endovasc Surg 2003; 25: 513–518.

38.

Weiss

Ramasastry

Eidelman

. Treatment of a cluster headache patient in a hyperbaric chamber. Headache 1989; 29: 109–110.

39.

Porta

Granella

Coppola

. Treatment of cluster headache attacks with hyperbaric oxygen. Cephalalgia 1991; 11(suppl 11): 236–237.

40.

Di Sabato

Fusco

Pelaia

. Hyperbaric oxygen therapy in cluster headache. Pain 1993; 52: 243–245.

41.

Nilsson Remahl

Ansjön

Lind

. Hyperbaric oxygen treatment of active cluster headache: a double-blind placebo-controlled cross-over study. Cephalalgia 2002; 22: 730–739.

42.

Pascual

Peralta

Sánchez

. Preventive effects of hyperbaric oxygen in cluster headache. Headache 1995; 35: 260–261.

43.

Nobre

Filho

Dominici

. Cluster headache associated with sleep apnoea. Cephalalgia 2003; 23: 276–279.

44.

Graff-Radford

Teruel

. Cluster headache and obstructive sleep apnea: are they related disorders? Curr Pain Headache Rep 2009; 13: 160–163.

45.

Zallek

Chervin

. Improvement in cluster headache after treatment for obstructive sleep apnea. Sleep Med 2000; 1: 135–138.

46.

Barloese

Jennum

Lund

. Sleep in cluster headache - beyond a temporal rapid eye movement relationship? Eur J Neurol 2015; 22: 656–e40.

47.

Demchenko

Welty-Wolf

Allen

. Similar but not the same: normobaric and hyperbaric pulmonary oxygen toxicity, the role of nitric oxide. Am J Physiol Lung Cell Mol Physiol 2007; 293: 229–238.

48.

Ashkenazi

Schwedt

. Cluster headache: acute and prophylactic therapy. Headache 2011; 51: 272–286.

49.

May

Leone

Afra

. EFNS guidelines on the treatment of cluster headache and other trigeminal-autonomic cephalalgias. Eur J Neurol 2006; 13: 1066–1077.

50.

Hojat

Mahdi

. Effect of different sitting posture on pulmonary function in students. J Physiol Pathophysiol 2011; 2: 29–33.

51.

Thys

Listro

Dozin

Marion

Rodenstein

. Determinants of FiO₂ with oxygen supplementation during noninvasive two-level positive pressure ventilation. Eur Respir J 2002; 19: 653–657.

52.

Steurer

Hoffmann

Dür

. Changes in arterial and transcutaneous oxygen and carbon dioxide tensions during and after voluntary hyperventilation. Respiration 1997; 64: 200–205.

53.

Ito

Kanno

Ibaraki

. Changes in human cerebral blood flow and cerebral blood volume during hypercapnia and hypocapnia measured by positron emission tomography. J Cereb Blood Flow Metab 2003; 23: 665–670.

54.

Hannerz

Jogestrand

. Provocation of unilateral pain in cluster headache patients by breathing CO₂. Headache 1995; 35: 38–43.

55.

Immink

Pott

Secher

. Hyperventilation, cerebral perfusion, and syncope. J Appl Physiol 2014; 116: 844–851.

56.

Kellogg

. In vivo mechanisms of cutaneous vasodilation and vasoconstriction in humans during thermoregulatory challenges. Journal of Applied Physiology 2006; 100: 1709–1718.

57.

Brown

Sanya

Hilz

. Effect of cold face stimulation on cerebral blood flow in humans. Brain Res Bull 2003; 61: 81–86.

58.

McLeod

Andrasik

Packard

. Cold room air inhalation to abort cluster headaches: an exploratory study. J Headache Pain 2002; 3: 33–36.

59.

Gross

Park

. Humidification of inspired gases during mechanical ventilation. Minerva Anestesiol 2012; 78: 496–502.

60.

Plafki

Peters

Almeling

. Complications and side effects of hyperbaric oxygen therapy. Aviat Space Environ Med 2000; 71: 119–124.

61.

Celli

MacNee

. ATS/RS Task Force. Standards for the diagnosis and treatment of patients with COPD: a summary of the ATS/ERS position paper. Eur Respir J 2004; 23: 932–946.

62.

Weinberger

Laskin

Heck

. Oxygen toxicity in premature infants. Toxicol Appl Pharmacol 2002; 181: 60–67.

63.

Mezu-Ndubuisi

Wanek

Chau

. Correspondence of retinal thinning and vasculopathy in mice with oxygen-induced retinopathy. Exp Eye Res 2014; 122: 119–122.

64.

Litt

Ziesche

Happak

. Burning HOT: revisiting guidelines associated with home oxygen therapy. Int J Burns Trauma 2012; 2: 167–170.

65.

Geerlings

Haane

Koehler

. Rebound following oxygen therapy in cluster headache. Cephalalgia 2011; 31: 1145–1149.

66.

Kudrow

. Association of sustained oxyhemoglobin desaturation and onset of cluster headache attacks. Headache 1990; 30: 474–480.

67.

Russell

. Cluster headache: severity and temporal profiles of attacks and patient activity prior to and during attacks. Cephalalgia 1981; 1: 209–216.

68.

Johnson

Ziemba

Garb

. Improvement in headaches with continuous positive airway pressure for obstructive sleep apnea: a retrospective analysis. Headache 2013; 53: 333–343.