Abstract
Background
Cluster headache is associated with structural abnormalities of the hypothalamus. We were interested in the association of cluster headache with endocrinological functional abnormalities. Therefore, we applied the apomorphine challenge test, which is a specific test of hypothalamic dopaminergic activation.
Methods
We enrolled 13 patients with cluster headache outside the bout and without medication. They were stimulated with 0.005 mg/kg of body weight subcutaneous apomorphine hydrochloride. After 45 and 60 minutes, growth hormone (GH), prolactin and cortisol were measured. The test was also applied to 14 sex- and age-matched healthy control subjects.
Results
There were significantly higher GH levels in healthy subjects as compared to cluster headache patients 45 minutes after injection (10.8 ± 10.8 versus 4.4 ± 7.4 ng/ml; p = 0.038). Only in cluster headache, the GH level after 60 minutes was not significantly different from the baseline. The levels of prolactin and cortisol did not show any significant differences between cluster headache patients and in healthy subjects.
Discussion
Our data suggest that cluster headache is associated with an impaired dopaminergic stimulation. This finding supports the body of evidence that cluster headache is associated with a functional abnormality of the hypothalamus and that this association is a primary (i.e. idiopathic) and not a secondary phenomenon during the bout.
Introduction
Cluster headache is associated with structural and functional abnormalities of the hypothalamus and/or structures nearby as shown by neuroimaging studies (1,2). In addition, prior studies on endocrinological functions also suggested an involvement of the hypothalamus in cluster headache (2–4). In these studies, however, the functional endocrinological stimulation of the hypothalamus in cluster headache patients has not been evaluated completely. It has also been debated whether the hypothalamic abnormalities in cluster headache are causally related to the disorder or are an unspecific epiphenomenon in general trigeminal pain processing (5).
The apomorphine challenge test, the first clinical application described in depression (6), has been developed as a method to study dopaminergic response and subsequently as a diagnostic tool for Parkinson’s disease (7–9). In particular, the growth hormone (GH) response as an indicator of dopaminergic stimulation is increased in Parkinson’s disease (7,10) but not in restless legs syndrome (11), although both conditions respond to dopaminergic medication. The excessive increase of GH after apomorphine stimulation in Parkinson’s disease is interpreted as an increased postsynaptic sensitivity of hypothalamic dopamine receptors in the course of a generalized degeneration of dopaminergic neurons. The response to apomorphine can also be considered as a functional hypothalamic response since the release of GH in the pituitary gland can be linked directly to the stimulation of the hypothalamus (12).
We were interested in the physiological responsiveness of the hypothalamus in cluster headache patients. Therefore, we used the apomorphine challenge test in cluster headache patients outside the bout and without medication to study the dopaminergic response and, thus, the functional responsiveness of the hypothalamus. Further, the results might elucidate the role of dopamine in cluster headache. As for control hormones, we also measured prolactin and cortisol, which are released by the pituitary gland but not by specific dopaminergic pathways.
Methods
Patients
We enrolled episodic cluster headache patients who were at least one month outside a bout. Their data were compared to the data of a healthy control group. Exclusion criteria were intake of any medication including prophylactic medication against cluster headache, smoking prior to the procedure, concomitant migraine, and any psychiatric or central neurological disorder. The study was approved by the local ethics committee. All patients were recruited from the headache outpatient clinic of the Department of Neurology, University of Münster. They gave written informed consent prior to the study procedure.
Study procedure
According to the protocol described previously (6,10), all patients received 0.005 mg/kg of body weight subcutaneous apomorphine hydrochloride (Teclapharm GmbH, Lüneburg, Germany) after fasting overnight. They were not pretreated with an antiemetic. All patients and subjects were examined in the morning at the same time of day. They were seated in a comfortable chair. After a rest of at least 30 minutes, a baseline blood sample was taken from an antecubital vein cannulated prior to resting.
According to Friess et al. (10), the maximal peaks of the GH increase occur around 45 to 60 minutes after injection of apomorphine. Therefore, GH, prolactin and cortisol were analyzed at baseline as well as 45 minutes and 60 minutes after subcutaneous apomorphine injection. The blood specimens were analyzed with commercially available radioimmunoassay kits at the Center for Laboratory Medicine, University of Münster, Germany.
Blood samples were collected from patients by venipuncture; for insertion of the needle, the arm was compressed, after that the compression was removed. Blood samples were centrifuged immediately after blood drawing. Sera were aliquoted, stored at −70℃ and analyzed in a routine hospital laboratory (Center for Laboratory Medicine, University Hospital Münster) using commercially available immunoassays. Concentrations of prolactin were determined using electrochemoluminescence immunoassay (ECLIA, Roche, Mannheim, Germany) on a Modular e170 automated analyzer. The intra-assay and interassay variabilities were <4.0% and <5.0%, respectively. Concentrations of cortisol and GH were determined using chemoluminescence immunoassays (CLIA, Siemens, Eschborn, Germany) on Advia Centaur or Immulite 2000 automated analyzers, respectively. The intra-assay and interassay variabilities were <3.7% and <5.4% for cortisol and <4.6% and <6.6% for GH. The analytical quality was constantly monitored according to regulations of the German Chamber of Physicians, and the laboratory took part in the external quality assessment schemes.
Statistical analysis
Data are presented as arithmetic mean with standard deviation or as percentage. We used non-parametric testing for group comparison, i.e. the Mann-Whitney-U-test for independent quantitative measures and the X2-test for qualitative measures. Significance level was set at p = 0.05.
Results
Demographic data of cluster headache patients and healthy control subjects presented as total number and as arithmetic mean with standard deviation. Statistical comparison by X2-test and Mann-Whitney-U-test.
ns: not significant.
Hormone levels of cluster headache patients and healthy control subjects at baseline (T0), 45 minutes (T45), and 60 minutes (T60) after injection of apomorphine presented as arithmetic mean with standard deviation. Statistical comparison between the groups was made by Mann-Whitney-U-test; statistical comparison between the three measures was made by Wilcoxon-test. The p level denotes the comparison between cluster headache patients and healthy control subjects.
p = 0.038 as compared to baseline; bp = 0.010 as compared to baseline; cp = 0.002 as compared to baseline; dp = 0.001 as compared to baseline; ep = 0.005 as compared to baseline; ns: not significant.
The Figure 1 illustrates the significantly different increase in steepness of GH levels in cluster headache patients and in healthy subjects.
Growth hormone (GH) levels in cluster headache patients and healthy control subjects during the study period (T0/T45/T60 denote baseline and 45 and 60 minutes after injection of apomorphine). The asterisk (*) denotes the significant difference in GH increase between the two groups (p = 0.038). For further statistical comparison and standard deviation, see Table 1.
There were no major side effects or adverse events caused by apomorphine. In the cluster headache and in the healthy subjects, nausea was reported by three and four subjects, respectively. Fatigue was reported by only one patient with cluster headache but by six healthy subjects. Yawning was exclusively reported by healthy subjects (n = five). However, no patient complained about severe gastrointestinal discomfort or unspecific headache after application of apomorphine.
Discussion
Our data suggest that cluster headache patients outside a bout show an impaired increase of GH after injection of apomorphine as compared to healthy control subjects. This finding can be interpreted as a decreased sensitivity of dopaminergic neurons in the hypothalamus in cluster headache patients. It is, thus, in concordance with the numerous findings of structural and functional abnormalities of the hypothalamus in cluster headache. Since prolactin and cortisol showed a similar response in cluster headache patients as in healthy control subjects, our finding seems to be specific for dopaminergic pathways. Further, this latter finding shows that the GH response is not just an unspecific stress reaction.
The validity of our data is also supported by the quantity of the GH response to apomorphine in healthy subjects, which is about the same as in older studies, and by the dissociation between GH and cortisol response that has also been described previously (13,14). Also, the mild decrease of prolactin after apomorphine injection had been described in the very first studies on this physiological mechanism (15). In another study, a reduced increase in GH between 24:00 h and 01:00 h in cluster headache patients in remission was observed, whereas nocturnal secretion of noradrenaline, cortisol and insulin did not differ significantly. This also indicates a permanent hypothalamic disturbance in cluster headache (16).
Interestingly, the sumatriptan-induced increase of GH levels was also impaired in patients with cluster headache during the bout; prolactin and adrenocorticotropic hormone (ACTH) responses to sumatriptan were also reduced in patients with cluster headache, both during and outside the bout. These findings suggest that both cerebral serotonergic functions mediated by 5-HT1D receptors and central dopaminergic functions are altered in patients with episodic cluster headache (17).
We examined the patients outside a bout and without any medication that could affect endocrinological responses in the central nervous system. Therefore, we believe that our findings represent a physiological pattern of cluster headache patients rather than an epiphenomenon of the pathophysiological state during a bout.
The apomorphine challenge test has been used in research on Parkinson’s disease. In this disorder, GH response to apomorphine stimulation is increased (9). In cluster headache, GH response to apomorphine was decreased possibly because of a downregulation of dopaminergic neurons. Also in migraine, dopaminergic response of the hypothalamus is impaired. Direct application of dopamine and dopamine receptor agonists onto trigeminocervical complex neurons inhibits their activation after nociceptive stimulation. In a clinical study, migraine patients showed a significantly higher incidence of dopaminergic symptoms such as yawning, nausea and sweating after apomorphine application, which was interpreted as a hypersensitivity to dopamine in migraine (18). The dopaminergic A11 nucleus of the hypothalamus has been identified as the likely source of this dopamine (19). Laboratory data suggest that the role of dopamine in migraine is more complex, perhaps because of the multiple receptors and levels of the brain involved in the disorder. These data suggest a reappraisal of dopaminergic therapeutic targets in migraine and cluster headache as our understanding of the role of this important biogenic amine is better characterized (19).
In an older study, no differences in the GH response were observed between cluster headache patients and healthy subjects when the insulin tolerance test, levodopa stimulation, and thyreotropin-releasing hormone stimulation were performed (20). However, since stimulation with apomorphine is more specific to dopaminergic neurons in the hypothalamus, this study does not contradict our findings. Furthermore, in the older studies on endocrinological response in cluster headache, patients were studied during and not outside the bout. Nonetheless, the finding of a lowered responsiveness of GH release after endocrinological stimulation (stimulated with GH-releasing hormone) in patients suffering from chronic migraine with medication overuse also highlights the relevance of the hypothalamic-pituitary-adrenal axis in headaches (21).
Interestingly, our finding of a lower sensitivity in cluster headache patients to dopaminergic stimulation is also supported by the side effects. Symptoms typical for dopaminergic stimulation (yawning, nausea, fatigue) were more often observed in the healthy subjects than in the cluster headache patients.
Our study has some limitations. First, the size of the study was rather small. Since we only enrolled patients outside the bout, it was difficult to convince patients to participate in a study in the morning with a subcutaneous drug application. Second, we did not control for a general decrease of dopaminergic neurons or dopamine receptors. In a consecutive study, it would be interesting to compare the stimulation results with the density of dopamine receptors as it has also been studied in Parkinson’s disease (22). It cannot be excluded, although it is unlikely from a clinical point of view, that cluster headache is associated with a degeneration of dopaminergic neurons, which would also explain our findings. Finally, it is possible that not a hypothalamic dysfunction but a dysfunction of the GH release in the pituitary gland is responsible for our finding. This is, however, also very unlikely since there is no evidence of other disturbances in the hormone release of the pituitary gland.
In conclusion, our findings support the body of evidence that cluster headache is associated with a functional abnormality of the hypothalamus and that this association is primary (i.e. idiopathic) and not a secondary phenomenon during the bout.
Clinical implications
Cluster headache is associated with low dopaminergic stimulation. Cluster headache shows functional impairment of the hypothalamus.
Footnotes
Funding
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
Conflict of interest
None declared.