Introduction

The intravenous glyceryl trinitate (GTN) headache model is well established [1, 2]. It causes a short-lasting headache in healthy volunteers and in migraine and cluster headache patients, the so-called immediate GTN headache [1, 2]. In addition, GTN causes a delayed headache in cluster headache [3] and migraine patients [4]. The immediate GTN headache is most likely due to a vasodilator effect of GTN on the extracerebral circulation [5] whereas the delayed headache/migraine is elicited by other mechanisms.

Administration of N-acetylcysteine, a sulfhydryl group donor, enhanced the headache-inducing effect of a GTN infusion in healthy volunteers [6]. This effect of N-acetylcysteine is most likely due to enhanced formation of NO from GTN.

In a recent Cochrane review [7] it was concluded that there was some evidence that hyperbaric oxygen was effective for the termination of acute migraine, and weak evidence that normobaric oxygen was similarly effective in cluster headache.

In the present study we tried to decrease the levels of NO in our GTN headache model by administering pure oxygen in a dose, 7 l/min which has been shown to be effective in the treatment of cluster headache in a double-blind trial [8, 9] and in clinical practice [8, 10]. Increased oxygen concentration most likely decreases NO concentration by oxidization of NO to NO3 and NO2.

Healthy volunteers were investigated in a crossover design.

Methods

Subjects

The study followed a randomized, double-blind, placebo-controlled, and cross-over design. The 2 study days were scheduled at least 7 days apart and included 18 healthy volunteers who had no prior history of migraine and who did not experience tension-type headache more than once a month. Their median age was 23 years (range 20–33 years), the male:female ratio was 1:1, and the median body weight 65 kg (range 52–86 kg). Prior to entry the subjects were challenged with 0.25 mg of GTN sublingually and were included only if they experienced headache. The participants gave informed consent prior to inclusion, and it was emphasized that they were free to withdraw at any time. The study was approved by The Ethics Committee of Copenhagen County.

Procedure

No caffeine containing beverages, alcohol, tobacco, or medication were allowed for the preceding 12 h prior to the study, and all had to be headache-free at start. After 30 min of rest, either pure oxygen or placebo (atmospheric air), 7 l/min, was inhaled for 35 min through a line fastened to an inflatable anesthesia mask. The gas bottles were hidden to the investigator and the participants. The proper gas was connected to the mask by a laboratory technician. GTN, 0.25 µg/kg/min, was infused for 20 min through a i.v. line in the right cubital vein using a volume-directed infusion pump (Braun Perfusor, Secura). Headache was rated on a verbal 0–10 numerical scale; 1 representing pre-pain, 5 a moderate headache, and 10 the worst possible headache. At pain scores above 0 the headache and accompanying symptoms were characterized according to a standardized questionnaire covering the IHS criteria for migraine without aura (IHS-classification, 12). The study ended 85 min after the start of the oxygen/placebo inhalation.

Statistics

Statistical analyses were performed with Statgraphic 7.0. Randomization and Wilcoxon–Pratt statistics were performed using Medstat, p < 0.05 was chosen as the statistical level of significance. Differences in the total, summed, and peak headache scores between the two study days were analyzed using Wilcoxon–Pratt’s test.

Results

All subjects completed the study. No serious adverse events were seen during or after the infusion of GTN.

Sixteen of 18 (89%) subjects experienced headache after the O2 inhalation and 17/18 (94%) on the day with air. The mean peak headache score in was 1.9 (O2) and 2.4 (air).

Oxygen did not significantly alter the summed headache scores or the peak headache scores (Wilcoxon–Pratt; summed p = 0.11, peak p = 0.25; Fig. 1).

Fig. 1
figure 1

Individual headache scores in the laboratory, 0–85 min after glyceryl trinitrate (with means in bold), n = 18. There was no difference between oxygen and air inhalation1

Full size image

The headache showed similar characteristics in the majority of the subjects. It was mild, bilateral, pressing, mainly fronto-temporal headache of short duration without accompanying symptoms. No subject fulfilled the diagnostic criteria for migraine without aura [11].

Discussion

The rationale for this study was the fact that increased oxygen tension will tend to increase the metabolism of NO. In an aqueous solution NO reacts rapidly and spontaneously with oxygen to produce mainly nitrite via the intermediary formation of nitrogen dioxide and dinitrogen trioxide [12].

A reduction in pO2 causes an increase in NO in some [13, 14] but not all [15, 16] studies. Theoretically, hypoxia could cause an increase in the effect of glyceryl trinitrate (GTN). This effect cannot be explored experimentally in humans outside a hypobaric chamber. In contrast, an increase in pO2 causes a shortening of the half-life of NO [17] due to increased metabolism and most likely a decrease in the effect of GTN. This possible effect of pure oxygen was explored in the present study.

The major result of the present study was that a therapeutic dose of oxygen inhalation (7 l/min) [8], did not reduce the headache induced by GTN infusion (0.25 µg/kg/min) in healthy volunteers.

Theoretically, this “negative” result could be due either to the fact that the GTN headache model is not sensitive enough to detect changes in NO levels or to a lack of oxygen to counteract the effect of GTN/NO.

That the GTN model is sensitive to changes in NO is demonstrated by the fact that a dose–response curve for the headache effect of GTN could be constructed [2] and by the fact that the sulfhydryl donor, N-acetylcysteine, increased the NTG-induced headache [6].

The dose of oxygen by inhalation was 7 l/min [8] but there was no effect on NTG-induced headache, most likely because the obtained increase of pO2 was not large enough to counteract the effect of GTN. Thus, in isolated guinea pig hearts increasing oxygen tension in aqueous solutions from 150 to 700 mmHg decreased NO half-life (5.2 s) by 32% [17]. Hundred percent oxygen inhalation in normal adults increases pO2 to about 600 mmHg [18]. Based on the in vitro in guinea pig study [17] this pO2 (600 mmHg) in humans may result in too small a decrease in NO levels to be detected in our headache model.

Our results suggest that the effect of pure oxygen in cluster headache may not be mediated via a reduction of NO.