Cerebrospinal fluid hypocretin-1 levels during the active period of cluster headache

Introduction

Cluster headache (CH) is a primary headache characterized by severe unilateral pain, ipsilateral autonomic symptoms and, in many cases, restlessness. A striking feature of CH is its diurnal and seasonal periodicity, suggesting that circadian and infradian rhythms regulate CH attacks.

The aetiology of CH is still unclear, although clinical, endocrine and neuroimaging studies implicate a role of the hypothalamus. In particular, the seasonal and circadian rhythmicity of the attacks points to the involvement of the suprachiasmatic nucleus, the master circadian clock of endogenous rhythms; further evidence comes from the close correlation with sleep and from the presence of cranial autonomic symptoms. The hypothalamic involvement is also supported by the efficacy of stereotactic deep brain stimulation of the posterior hypothalamic area in chronic drug-resistant CH patients (1).

Hypocretins (orexins) are hypothalamic neuropeptides involved in a wide range of physiological processes in mammals, such as feeding, sleep–wake regulation, autonomic function, reward and drug addiction, and central pain modulation (2). A link between these neuropeptides and nociceptive or autonomic phenomena observed in primary headaches has recently been suggested (3,4). Genetic studies in humans confirmed a role for the hypocretinergic system in CH. In particular, the G1246A polymorphism of the hypocretin receptor 2 gene has been associated with CH (5), a remarkable association because hypocretin-containing cells are located exclusively in the posterolateral hypothalamus.

We tested cerebrospinal fluid (CSF) hypocretin-1 (orexin-A) levels in 10 CH patients during an active cluster period.

Patients and methods

In 10 consecutive unrelated Italian patients (nine men and one woman, mean age 41 ± 12.6 years, range 28–70) with CH (six episodic and four chronic CH) we assessed disease duration, the elapsed time from the previous attack and from the onset of the current active cluster period (Table 1). Brain MRI and neurological examination of all patients were unremarkable. Migraine and sleep disorders were excluded in all patients. All patients gave informed consent to perform a lumbar puncture (LP) and were diagnosed according to ICHD-II criteria by a neurologist expert in headache disorders (6). LPs were performed using atraumatic needles between 08.00 and 11.00 with the patient in the seated position. CSF was immediately frozen at −80°C. CSF hypocretin-1 levels were determined in all patients in a single radioimmunoassay, as previously described (7). For quantitative data the Mann–Whitney U-test was used. Significance level was set at p = 0.05. The Institutional Review Board of the Department of Neurological Sciences of the University of Bologna Medical School approved the project. Normal values (> 320 pg/ml) were ascertained by determining CSF hypocretin-1 levels in 20 healthy subjects (11 males and nine females, mean age 44 years, range 17–79), as previously described (mean 497 pg/ml; range 350–603) (7). It is well accepted that ethnicity, age and gender do not influence hypocretin-1 levels (8).

OPEN IN VIEWER

Table 1.

Patients’ clinical features and CSF hypocretin-1 levels

Patient no. Diagnosis (E/C)*	Sex	Age, years	Age at CH onset, years	Attacks/ day	Time from last attack and LP (h.min)	Time from the onset of last cluster period and LP	Therapy	CSF hypocretin-1 (pg/ml) NV > 320 pg/ml
1. E	M	28	23	3	8.15	4 months	Verapamil	451
							Ergotamine
							O2
2. E	M	47	27	1	6.20	3 months	Verapamil	304
							Sumatriptan
3. E	M	33	16	3	2.05	7 days	Lithium	386
							Methylprednisolone
							Sumatriptan
4. E	M	30	13	1 or 2	8.00	15 days	Verapamil	546
							Prednisone
							Sumatriptan
5. E	M	55	32	1	22.00	3 days	Verapamil	350
							Sumatriptan
6. E	M	34	31	2 or 3	4.00	5 months	Verapamil	506
							Lithium
							Sumatriptan
7. C	M	55	17	3 or 4	18.10	NA	Methylprednisolone	639
							Sumatriptan
8. C	F	30	16	5	3.10	NA	Verapamil	546
							Methysergide
							Indometacin
							Sumatriptan
9. C	M	28	23	2	22.00	5 years	Verapamil	475
							Methysergide
							Lithium
							Sumatriptan
10. C	M	70	50	4	11.00	14 years	Methylprednisolone	370
							Duloxetina
							Pregabalin
							O2

Results

CSF hypocretin-1 levels were within the normal range (mean 457.3 ± 104.98 pg/ml, range 304–639; p = 0.3) in our 10 patients, with a slight reduction in one case (304 pg/ml, but still > 200 pg/ml). There were no associations between CSF hypocretin-1 levels and the clinical features of CH (duration of the disease, drugs assumed, time elapsed from the last attack, number of attacks per day and circadian rhythmicity of the attacks) (Table 1) (Figure 1). A trend towards higher hypocretin-1 levels was disclosed in patients with chronic CH compared to episodic CH (mean ± SD: 507.5 ± 113.63 vs. 457.3 ± 104.98; p = 0.24), mainly because of the presence of a high hypocretin-1 level in one chronic CH patient (Table 1: patient 7). OPEN IN VIEWER

Discussion

The aim of this study was to test the hypothesis that CSF hypocretin-1 levels in CH patients during an active cluster period is disturbed as compared to healthy controls. We detected normal CSF hypocretin-1 levels in nine patients, with a slight reduction in one case. A trend towards higher hypocretin-1 levels was disclosed in patients with chronic CH compared to episodic CH but the difference was not significant. Higher hypocretin-1 levels were previously reported in patients affected by chronic migraine and medication overuse headache compared to healthy controls (9), suggesting an association between chronic head pain and hypocretin-1 signalling. Thus CSF hypocretin-1 levels seem not to be significantly altered regardless of the clinical features of CH. However, based on peripheral CSF levels, we cannot completely exclude a functional involvement of the hypothalamic hypocretinergic system in the pathogenesis of CH. Because a partial loss of hypocretin-producing neurons is not necessarily followed by a clear-cut reduction of CSF hypocretin-1 concentration (10), future studies are needed to evaluate hypocretin neuronal defect. Moreover, our preliminary study, although the sample was relatively small, found a relatively low hypocretin-1 concentration during the active period of episodic CH in one patient. A two-fold decrease in hypocretin-1 levels between the asymptomatic interval and the symptomatic period was previously detected in Kleine–Levin syndrome, despite the fact that the lower hypocretin-1 concentration was still within the normal range (11). Given that CH, like Kleine–Levin syndrome, is an episodic brain disorder, additional measurements of hypocretin-1 in both a symptomatic period and an asymptomatic interval are warranted.