Abstract
Clinically, cluster headache (CH) is characterized by stereotypical, strictly unilateral, very severe pain attacks, lasting 15 min to 3 h. These attacks are usually accompanied by ipsilateral autonomic facial features, such as miosis, conjunctival injection and lacrimation. Although recent findings point towards the hypothalamus as a key structure in the pathophysiology of CH (1), the mechanisms of this primary headache type are far from being understood. A number of studies have previously described hormonal influences, but not progesterone, in this headache type (2).
We report a woman suffering from CH, recurrently triggered by hormone treatment associated with in vitro fertilization (IVF). The close temporal relationship between progesterone treatment and CH attacks in our patient might be of pathophysiological interest.
Case report
History
A 34-year-old female health professional sought infertility treatment due to severe endometriosis and subfertiliy of her husband. Two cycles of intracytoplasmatic sperm injection (ICSI) therapy were performed. The patient reported the onset of serial headache attacks during high-dose hormonal treatment in the context of her treatment.
Her medical history was unremarkable except for hypothyroidism, treated with levothyroxine. She denied alcohol consumption or smoking. Her mother had had migraine without aura. The family history for CH was negative.
Headache description
The onset of her headache was rapid, with the intensity reaching its maximum (10 out of 10 on a verbal rating scale) within 1 min. Pain was strictly right-sided, with periorbital localization and described as ‘a knife behind the eye’. The attacks occurred mostly at night and almost always between midnight and 02.00 h. Maximum headache frequency was six per day, then without a circadian pattern. Within the two treatment cycles, the first attacks lasted for approximately 30 min, and within a few days increased to a maximum of 1.5 h. Attacks were accompanied by lacrimation and nasal congestion, and miosis was observed during one single attack. Nausea, phono- or photophobia, premonitory symptoms or other autonomic features were not present. During headache attacks, the patient was restless and walked around. Riding a home bicycle trainer or jogging provided some relief. Paracetamol did not influence the pain. The patient refused other headache medications in the context of the IVF programme.
The only previous headaches the patient was able to recall had occurred during a safari some time ago. She reported strictly right-sided, periorbital, severe and lancinating pain with a duration of approximately 15 min. She did not recall any autonomic features or restlessness.
Clinical and laboratory findings
Neurological examination as well as magnetic resonance imaging of the brain were unremarkable. Before the ICSI treatment was started, blood levels of follicle-stimulating hormone (FSH), luteinizing hormone (LH), prolactin, thyroid-stimulating hormone, basal and free thyroxin had been normal. When pregnancy was confirmed, laboratory examination revealed mild hypothyroidism, and the levotyroxine dose was adjusted accordingly.
Course of hormonal treatment and of headache
For the first of two ICSI treatments, a modified short protocol of the clinic for reproductive endocrinology, University Hospital Zurich, was used. First, down-regulation of the natural cycle was initiated using norethisterone 10 mg/day for 18 days, followed by stimulation with recombinant FSH (200 IU for 3 days, then 300 IU for 7 days) (3). A gonadotropin-releasing hormone (GnRH) analogue was added daily. Thirty-five hours before oocyte recovery, on day 13 of the stimulation, 6500 IE recombinant human chorionic gonadotrophin (HCG) was injected for final ovarian maturation. After the transfer of two embryos, 90 mg progesterone was administered intravaginally once a day, for 2 weeks, for luteal phase support. When the pregnancy test was confirmed negative, the progesterone was stopped.
The patient did not have any headaches during the stimulation periods. Only when progesterone treatment began did the headaches occur. After discontinuation of progesterone substitution, she became headache free each time.
During the second treatment, cryoconserved zygotes were transferred into a natural cycle. The thawing cycle was repeated in an artifical cycle (synchronized with estradiole 3× 2 g). After each transfer, intravaginal progesterone was daily administered as follows: 90 mg in the morning, within a natural cycle; within an artificial cycle, 400 mg in the morning, 200 mg at noon and 400 mg in the evening. If the pregnancy test was confirmed negative 2 weeks after each transfer, progesterone was discontinued.
During each of the four progesterone administration periods, similar headaches recurred. The same headache type recurred also during the second ICSI therapy, when a long stimulation protocol was followed. For down-regulation, the GnRH analogue was injected daily. After 23 days a human gonadotropin (LH/FSH) was added (225 IE for 11 days). HCG was given again and, following the embryo transfer, intravaginal supplementation with progesterone 3× 200 mg daily was started. The patient became pregnant this time and progesterone treatment with 3× 200 mg daily was continued for 10 weeks of gestation, during which time the patient suffered severe headache attacks. After 5 weeks of gestation the dose of the progesterone treatment was reduced to 2× 200 mg daily, without any improvement of headaches.
After 10 weeks of gestation, progesterone intake was discontinued as sufficient progesterone synthesis by the placenta was expected. The headache attacks gradually diminished in intensity and disappeared within a week. Thereafter there were no recurrences.
Discussion
Headache attacks as described by the patient fulfil the criteria of the International Headache Society for episodic cluster headache (ICHD-II 3.1), including localization, duration, the presence of autonomic features, and restlessness (4). Formally, criteria for ICHD-II 8.3.1 (exogenous hormone-induced headache) are fulfilled (4), also, as in our patient the headaches developed in close temporal association with hormone therapy, whereas they discontinued when therapy was stopped.
Exogenous hormone-induced headache is a diagnostic category which plays a role mostly in the context of hormonal contraception or hormone replacement therapy after menopause using oestrogens, and, importantly, the headaches triggered by oestrogens mostly fulfil migraine criteria. In our patient, the clinical situation was entirely different as the headaches fulfil CH criteria and the hormone treatment is progesterone, as part of IVF.
The close temporal relationship between several courses of progesterone treatment and CH attacks in our patient implies a causal relationship.
There is evidence for a prominent role of the hypothalamus in the pathophysiology of CH (1). The central role of the hypothalamus is supported by the circadian and circannual pattern of CH, several imaging studies (5, 6) and, finally, by the therapeutic success with hypothalamic brain stimulation (7). In CH patients, alterations in the plasma levels of cortisol, testosterone, gonadotropins, prolactin, growth hormone and thyreotropin (2, 8) have been found, all of which are controlled by hypothalamic feedback loops.
In our patient FSH and LH were used. Progesterone supplementation in IVF is required to increase the probability of successful implantation. The recurrence of cluster attacks during progesterone therapy and the remissions at the end of each treatment period make a causative role highly probable. The longest and most frequent attacks occurred after successful IVF. This can be explained by the effects of placental progesterone production during pregnancy in addition to therapeutic progesterone administration over several weeks. Importantly in this context, ovarian hormones passively diffuse through the blood–brain barrier (9).
The pathophysiological aspect of most interest could be the thermogenic effects of progesterone (10). The changes in body temperature of about 0.5°C during the luteal phase of a normal menstrual cycle are attributed to this effect (10). A study of trigger factors in 200 cluster patients revealed increased body temperature caused by either a hot environment or exertion as a frequent but rarely considered trigger (11). The previously reported attacks during a safari could also be seen in this context.
During pregnancy, latent hypothyroidism was diagnosed. Interestingly, a pathological thyreotropin releasing hormone test had been reported in cluster patients during pain episodes, but not during remission (12).
The question of how to treat CH during pregnancy was difficult, because valid data on this topic are lacking. The patient, a health professional herself, decided against all therapeutic options except paracetamol.
In conclusion, to our knowledge this is the first reported patient suffering from CH triggered by hormonal stimulation with progesterone during in vitro fertilization. Additional studies on progesterone levels in cluster patients are needed to support our findings in this single case study. Better understanding of the hormonal aspects of CH is required.
Acknowledgements
C.G. was research fellow at the University of Zürich at the time of the preparation of this work. He was supported by an MSD headache research grant.