Abstract
Changes in intraocular pressure (IOP) elicited by a Valsalva manoeuvre were studied in 11 male patients (mean age 39.8 years) suffering from episodic cluster headache (CH), and 12 healthy male controls (mean age 39.9 years). The tests were performed at rest and while exhaling hard through a mouthpiece connected to a mercury manometer. In the CH group, during symptomatic periods, between attacks, Valsalva manoeuvre elicited an asymmetric increase in IOP with significantly higher values on the symptomatic side (P = 0011), whereas no asymmetric increments in IOP were found during asymptomatic periods. Outside the cluster period the IOP values both baseline and with Valsalva manoeuvre did not differ from controls. The increment in IOP took place within a few seconds, as in spontaneous CH attacks, thus pointing to a rapid increase in intraocular blood volume or vasodilatation. These findings may reflect a latent increased vascular reactivity of the symptomatic orbit during CH period.
Introduction
Cluster headache (CH) attacks are accompanied by bilateral increase in intraocular pressure (IOP) predominantly on the eye on the symptomatic side (1). The increase in IOP during CH attacks is so rapid that it is attributed to an ictal blood volume into the orbit rather than to changes in dynamics of aqueous humor, which require more time.
In CH, an abnormal vascular reactivity on the symptomatic orbit may be latent throughout all the cluster period. We have endeavoured to test whether the asymmetry in IOP observed during CH attacks could be reproduced interictally by a provocative test, such as Valsalva manoeuvre which initially leads to both a transient increase in blood pressure and a reduction in venous return to the heart. This approach provides easily obtained information on the vascular reactivity of the orbital area that may be of clinical importance in CH.
Materials and methods
In a 16-month period, we enrolled 11 consecutive episodic CH patients starting a symptomatic cluster period. Subjects were recruited sequentially from the Headache Programme of the Fundación Hospital Alcorcón. All patients had a strictly unilateral CH. Twelve healthy age- and gender-matched volunteers were used as controls. Diagnosis was made following International Headache Society criteria (2). Informed consent was obtained according to the Fundación Hospital Alcorcón Ethics Committee. None of the patients had started preventive medication or taken abortive drugs 24 h before the testing. No other treatment 24 h before testing was allowed. All patients had a normal neurology and ophthalmology exams as well as a brain magnetic resonance imaging (standard protocol).
All CH patients were tested interictally during a symptomatic period. In addition, CH patients were tested during asymptomatic periods. IOP was measured using the Perkins aplanation tonometer (3) (MK2, CE-0120; Clement Clark Int., Harlow, Essex, UK) under the following conditions: rest (baseline measurement), and during three consecutive Valsalva manoeuvres (an average of the three values was used). A concordance of < 10% variation between the three values per eye was verified in all the patients and controls. The IOP measurement required the instillation of an anaesthetic eye drop. Special attention was paid to the presence of any autonomic symptoms and pain during the testing and the following hours.
Measurements were made with all the subjects seated. For the Valsalva manoeuvre patients and controls were instructed to blow through a tube connected to a manometer until 40 mmHg was reached, and to maintain that pressure for 10 s. In order to ensure a constant pressure, the manometer was designed with a valve that opened when pressures exceeded 40 mmHg. Therefore, a similar effort was achieved by all the patients and controls. The ophthalmologist in charge of measuring the IOP was blinded regarding both the diagnosis and the symptomatic side. Each IOP measurement was performed during Valsalva phase 2, a few seconds after starting blowing and before release of intrathoracic pressure, following the same sequence of first right eye followed by the left.
A normal distribution of the sample was assessed and verified by Kolmogorov–Smirnov and Shapiro–Wilk contrasts. Both Student's t- and non-parametric Wilcoxon tests were applied (statistical package SPSS version 9). Data are expressed as mean ± SD. Statistical significance was assigned at the 5% level.
Results
A total of 11 episodic CH male patients, mean age 39.8 ± 11.6 years (range 28–59) with 2.5 ± 0.8 years of cluster headache history, were studied. The 12 male healthy volunteers were within the same age range as study subjects (39.9 ± 10.3 years).
Patients were studied an average of 16.5 ± 5.3 h after the last attack. Between attacks baseline IOP was similar in both eyes and within normal limits (14.5 ± 1.8 mmHg asymptomatic side, 14.4 ± 2.3 mmHg: symptomatic side) (see Table 1). During the Valsalva manoeuvre there was a significant asymmetric increase in IOP: the increase in IOP reached significant levels vs. baseline on the symptomatic side, as well as compared with the contralateral side (P = 0.011, Wilcoxon test). The Valsalva manoeuvre did not elicit either pain or autonomic changes characteristic of the CH attack. Outside the cluster period the IOP values both baseline and with Valsalva manoeuvre did not differ from controls (Table 1).
Intraocular pressure (mean ± SD, mmHg) elicited by Valsalva manoeuvre
†In healthy volunteers the symptomatic side is arbitrarily assigned to the right eye and the asymptomatic to the left eye.
∗P = 0.011 (non-parametric Wilcoxon test) and P = 0.005 (Student's t-test).
Discussion
Valsalva manoeuvre is used to stimulate and investigate the response in the ocular vascular system and in the autonomic nervous system (4). The following factors determine the level of IOP: rate of aqueous secretion, resistance of aqueous outflow and episcleral venous pressure. In normal subjects it has been shown that during Valsalva the IOP may increase (5, 6).
In our patients, Valsalva manoeuvre elicited an asymmetric increase in IOP with significantly higher values on the symptomatic side during the CH symptomatic period, between attacks. The increment in IOP took place within a few seconds, as in spontaneous CH attacks, thus pointing to a quick increase in intraocular blood volume.
Both increase in episcleral pressure and engorgement of the choroidal vessels may also produce an increase in the total ocular volume (7). Theoretically, a sudden vasodilatation of the ocular arteries or an abrupt reduction in the ocular venous return to the cavernous sinus may account for the precipitated changes in IOP. This suggests that in CH the symptomatic orbital area has to accommodate higher blood volume than the non-symptomatic orbital area.
The test did not elicit either pain or autonomic changes, suggesting those vascular changes alone are not sufficient to justify the clinical picture and that specific parasympathetic activation is required for the conjunctival hyperaemia.
Contraction of the extraocular muscles during the procedure may contribute to increased IOP. Since an asymmetrical simultaneous co-contraction of the extraocular muscles is virtually impossible, we excluded that such a possibility could have substantially influenced our results. On the other hand, our control group did not show asymmetries in IOP during the procedure. Indeed, the unilateral preponderance of increased IOP makes unlikely both intracranial or a systemic cardiovascular disorder as responsible for Valsalva-precipitated asymmetrical increase in IOP.
In conclusion, these findings indicate that throughout the symptomatic periods of CH there is either an increased reactivity of the orbital and eye vascular bed or an impaired venous blood drainage to the cavernous sinus (8). Further studies with continuous monitoring of the IOP and ocular eco-Doppler are underway.
Footnotes
Acknowledgements
The work described here was supported by Astra Zeneca (Spain). M.S.d.R. is supported by PPA Humphrey-IHS-GlaxoSmithkline 2000 Research Fellowship.