Ecacy and safety of oral zolmitriptan in the preictal phase of episodic cluster headache

Recent studies on cluster headache suggested that headache includes three phases.Oral zolmitriptan is effective in the treatment of acute headache attack.We aim to describe the clinical characteristics of different phases of episodic cluster headache(eCH) and observe the ecacy and safety of oral zolmitriptan in preictal phase of eCH attacks. Forty-eight patients with eCH were enrolled in this study. We divided the CH attack into three phases: preictal, ictal and postictal phase and analyzed the clinical features of each phase. We collected the data of twenty-three patients who took zolmitriptan in the preictal phase of three consecutive CH attacks. The Visual Analogue Scale (VAS) score and duration of headache after three consecutive treatments were compared with the baseline respectively. Chi square test, t test, Mann Whitney U test, Pearson linear correlation analysis, paired t-test and paired non-parametric rank-sum test were used for statistical analysis.

Recent studies on cluster headache suggested that headache includes three phases.Oral zolmitriptan is effective in the treatment of acute headache attack.We aim to describe the clinical characteristics of different phases of episodic cluster headache(eCH) and observe the e cacy and safety of oral zolmitriptan in preictal phase of eCH attacks.

Methods
Forty-eight patients with eCH were enrolled in this study. We divided the CH attack into three phases: preictal, ictal and postictal phase and analyzed the clinical features of each phase. We collected the data of twenty-three patients who took zolmitriptan in the preictal phase of three consecutive CH attacks. The Visual Analogue Scale (VAS) score and duration of headache after three consecutive treatments were compared with the baseline respectively. Chi square test, t test, Mann Whitney U test, Pearson linear correlation analysis, paired t-test and paired non-parametric rank-sum test were used for statistical analysis.

Results
Forty-ve patients (93.8%) reported the presence of preictal symptoms. The most frequent preictal symptom was mild to moderate head discomfort in the area of the subsequent serious headache, accounting for 87.5%. The duration of preictal period was positively correlated with the duration of headache. The Pearson correlation coe cient was 0.313 (P=0.018). Among the twenty-three patients taking oral zolmitriptan in the preictal phase, the VAS max score was signi cantly lower than that of baseline headache (P<0.01). The duration of headache after three preictal administration was signi cantly shorter than that of baseline headache (P<0.01). The incidence of adverse reactions after taking zolmitriptan was 10.1%, including head drowsiness, neck stiffness and mild chest tightness.

Conclusions
Preictal symptoms were common in CH. Mild to moderate discomfort in the subsequent serious headache was the most common symptom in the preictal phase. Oral administration of zolmitriptan in the preictal period could reduce the severity of headache and shorten the duration of headache in a cluster period without serious adverse reactions.

Background
Cluster headache is a trigeminal autonomic cephalalgia which is characterized by strictly unilateral headache attacks lasting 15-180 minutes and occurring up to eight times a day accompanied by at least one autonomic symptom ipsilateral to the pain or a sense of agitation, or both [1]. It can be divided into episodic cluster headache (eCH) and chronic cluster headache (cCH). ECH accounts for 90% of CH [2].
Patients with eCH have at least two cluster periods lasting for 7 days to 1 year without treatment, and the remission period is ≥ 1 month. Owing to the excruciating and highly-frequent recurrence during the cluster bout, CH has been referred to as the "suicide headache". 15.9% of Asian patients suffering from acute headache attack had suicidal ideation, while the suicidal ideation decreased signi cantly during headache intermission [3]. The recurrent intolerable attacks of the disorder seriously affected the patient's work and life [4] and caused heavy social and economic burden [5]. Recent studies have found that one attack of CH did not just include pain phase. Preictal symptoms were very frequent in CH [6]. Given that these symptoms occur relatively early in the preictal phase and are frequent, there could be a potential bene t if starting abortive treatment when these symptoms occur. While nowadays the treatment of acute attack has been mainly conducted in the ictal phase, in which the attack had already made a serious impact on patient. Studies found that the clinical bene ts were more pronounced when the patients took medicine at the initial period of migraine attack than delayed medication [7][8]. The pathophysiology of CH was similar to that of migraine to some extent, so we speculated that administration in preictal phase of CH could effectively relieve or terminate the headache attack as soon as possible. We aimed to investigate the clinical characteristics of the preictal phase of eCH in our headache clinic and to observe the e cacy and safety of oral zolmitriptan in the preictal phase of eCH patients within the cluster bout.

Study design and Participants
According to ICHD-3 ( C. Either or both of the following: 1. at least one of the following symptoms or signs, ipsilateral to the headache: All patients completed a structured questionnaire designed for the assessment of eCH. We collected clinical information about current and previous bouts, including age, gender, family history of headache, course of disease, location, manifestation of headache (tabbing pain, dull sensation, distending pain and throbbing pain), headache duration, frequency, severity (VAS score, numeric rating scale from 0 to 10: 0 meant painless; 1-3 mild pain; 4-6 moderate pain; 7-8 severe pain; 9-10 extremely severe pain) [9], diurnal rhythm, inducements, phases and accompanying symptoms.
We collected the clinical data of 23 patients who took zolmitriptan during three consecutive preictal phases, including the baseline clinical data, medication time, VAS scores at 15 minutes, 30 minutes and 60 minutes after treatment, VAS max score, duration of headache and adverse reactions. The VAS score and duration after three consecutive medications were compared with that of the baseline, respectively.
Zolmitriptan was produced by wante Pharmaceutical (Hainan) Co., Ltd., and the trade name is zolmitriptan tablets (kaqu), national drug approval word H20061077. The dose was 2.5mg/tablet. The administration in preictal phase de ned as oral administration of zolmitriptan within 15 minutes of preictal phase[8].

Phases and accompanying symptoms
We divided the phases into three sections: preictal phase, ictal phase and postictal phase. The ictal phase was de ned according to the ICHD-3 as the period of severe or extremely severe periorbital and/or temporal pain with accompanying symptoms or restlessness. The preictal phase was de ned as the time that pain symptoms, autonomic nerve symptoms or other symptoms began to appear until severe or very serious pain occurred. The postictal phase was the time that severe pain got into mild until it disappeared and other accompanied symptoms disappeared completely.
All symptoms were grouped into 3 categories: local and painful symptoms (comprising local painful sensations in the subsequent severe headache, fullness of the ear, neck pain); local and painless symptoms and signs (conjunctival injection, lacrimation, rhinorrhea, nasal congestion, forehead facial sweating ptosis or miosis) and general symptoms (di culty concentrating, restlessness, mood changes, etc).

Data comparison
The eCH patients with preictal phase were divided into two groups, according to the duration of preictal phase, ≤15 min group and > 15 min group. The number of symptoms, headache severity in the ictal phase and headache duration were compared between the two groups. For 23 eCH patients treated within preictal phase, the VAS score and duration after three-time treatments were compared with that of the baseline respectively.

Statistical analyses
Statistical analyses were performed using SPSS23.0 software. Chi square test was used for categorical variables, which were presented with the number of cases (%). T-test was used for continuous variables in accordance with normal distribution, which were presented with mean ± standard deviation. The metrological data that did not conform to the normal distribution were represented by the median (quartile range), using Mann-Whitney U test. Pearson linear correlation analysis was used for correlation analysis. P < 0.05 was statistically signi cant.
The VAS score before and after treatment was presented with medians and interquartile ranges (IQR). The duration was presented with mean and standard deviation. The duration of headache after treatments in the preictal phase for three times was respectively compared with that at baseline by paired t-test, and VAS scores by paired non-parametric rank-sum test, and the test level was adjusted to 0.017 (0.05/3).

Results
Demographic and clinical characteristics of eCH A total of 48 eCH patients were recruited in the study. The male to female ratio was 5:  (Figure 1). Clinical characteristics of the preictal phase of eCH Among the 48 eCH patients included in our study, 45 patients (93.8%) reported the presence of preictal symptoms. The most frequent preictal symptom was mild to moderate head discomfort in the area of the subsequent attack, accounting for 87.5%. Forehead and facial sweating was reported in 9 patients (18.7%), lacrimation in 7 (14.5%) and nasal congestion in 6 (12.5%) during the preictal phase. The frequent general symptoms were mood changes (33.3%), di culty in concentration (31.2%) and restlessness (31.2%).

Group comparisons
Among the 48 eCH patients, 45 had preictal phase while 3 patients reported no obvious preictal phase. The preictal phase duration ranged from 5 to 60 min, of which 5-30 min was the most common (71%).
( Figure 2). The duration of headache in the group with preictal phase > 15 min was longer than that ≤ 15 min, and the difference between two groups was statistically signi cant (P = 0.018)( Table 3). The duration of preictal phase was positively correlated with headache duration (Pearson correlation coe cient was 0.313). There was no correlation on the number of symptoms and VAS score between the two groups. The median score of VAS max after three medications were all 5, while the baseline was 9. The duration of attack after three times administration of zolmitriptan was 48.2 ± 24.5 minutes, 45.7 ± 21.8 minutes, 52.6 ±31.9 minutes, respectively, while the baseline was 75.7±38.1 minutes. The differences in VAS max scores and duration before and after treatment were statistically signi cant (P<0.01) ( Table 4). The proportions of mild or painless headache at 30 minutes after each treatment were 56%, 70% and 61%, respectively. No difference was found among the three groups. After three-time preictal administration, the median VAS scores were 3,3,2, respectively. Scores at 15 minutes were 5,4,4, 30 minutes were 2,1,1 and 60 minutes were all 0 ( Figure 3).

Adverse reactions
For the rst treatment, two patients felt dizzy and one patient reported dizziness and mild chest tightness. One patient was accompanied with drowsiness and two patients with dizziness and neck stiffness for twice. The incidence of adverse reactions in our study was 10.1%, including dizziness, neck stiffness, drowsiness and mild chest tightness, all of which were relatively mild and could be relieved spontaneously.

Discussion
We conducted a single-center observational study to investigate the clinical characteristics of preictal phase of eCH and evaluate the e cacy and safety of oral zolmitriptan in the preictal phase. Our main ndings were as follows: preictal symptoms were observed in most of the eCH patients, and the preictal duration was positively correlated with the headache duration within the current bouts. Preictal administration of zolmitriptan in eCH patients could signi cantly shorten headache duration, reduce the headache severity, and prevent progression of severe headache.
In our study, preictal symptoms occurred in forty-ve (93.8%) patients. The duration of preictal phase was less than 1 hour, usually lasting for 5 to 30 minutes, which was in line with the study in Korean [10] and Rossi study [11], signi cantly shorter than the duration of migraine preictal symptoms. Mild to moderate discomfort in the subsequent serious headache was the most common preictal symptom, accounting for 87.5%, manifesting as dull, stabbing and distending sensation. Blau [12] and Engel [13] were the rst to investigate pre-attack symptoms of CH. They de ned the local and painful symptoms as "pain accumulation" of the headache. The occurrence of neck pain in preictal phase accounted for 4.1%, which indicated that the disfunction of trigeminal cervical complex and hypothalamus may be involved in the pathophysiology of CH preictal phase. Among the autonomic symptoms, conjunctival congestion and lacrimation were observed in 10% of CH patients, and the incidence of those symptoms was signi cantly lower than the 40% reported in the Western countries [13], similar to the incidence of 11% reported by Soohyun Cho in Korea [10]. None of the patients in this study reported miosis in the preictal phase, probably because this is di cult to determine if not paid speci c attention.
Mood changes and di culty in concentration occurred in 33.3% and 31.2% of our patients, respectively. Previous neuroimaging research showed that the frontal cortex may be involved in the pathophysiology of the preictal phase. MR diffusion tensor imaging study demonstrated that the white matter connection of different hypothalamic nuclei showed increasing trend of connection between right frontal and cerebral hemisphere [14]. Changes in mood and attention suggested the involvement of limbic system. While yawning and physical energy decrease may be mediated by the central dopaminergic system [15][16].In the study of Laurell, about one third of migraine patients had preictal symptoms such as yawning, mood changes, restlessness and photophobia, which directly or indirectly indicated that the hypothalamus was involved in the pathophysiology of migraine [17].The recent neuroimaging studies of migraine have con rmed that hypothalamic activation was involved in the occurrence of migraine preictal symptoms. Alterations in the functional connections between the hypothalamus, the trigeminal nucleus and the rostral pons promoted the procession of migraine [18][19][20][21]. It was well known that hypothalamus and trigeminal nerve system were involved in the pathophysiological mechanism of CH, therefore the preictal symptoms may be related to the activation of these sites [20]. Previous studies showed that the trigeminal neurovascular system, parasympathetic bers, and hypothalamus took part in the onset of CH attack. Soohyun Cho speculated that these above structures may be involved in the early phase before the onset of CH [10]. The most recent study of CH proposed the theory of"shadow attacks", which referred to the low-level self-limiting ipsilateral pain that occurred preceding the rst episodic attack in the upcoming CH bout [22][23]. The investigators considered the symptoms as predictive sign for cluster bout [24]. Further dynamic functional imaging studies on the "shadow attacks" could be helpful to reveal the pathophysiology of the disease.
In our study, we found that the duration of preictal phase was positively correlated with the headache duration (Pearson's correlation coe cient was 0.313) while the numbers of symptoms and the severity of headache had no correlation with the duration of preictal phase. Combined with the previous studies of CH, it was known that the plasma concentrations of Calcitonin gene-related peptide (CGRP) and Vasoactive intestinal peptide (VIP) increased during attacks [25]. Studied on experimental animal models demonstrated that trigeminal nerve stimulation could lead to the release of CGRP, VIP, Pituitary adenylate cyclase activating polypeptide (PACAP) and other in ammatory peptides. PACAP could induce the activation of related neurons (such as superior salivatory nucleus(SSN), sphenopalatine ganglion(SPG) and trigeminal ganglion (TG), and drive the onset of typical symptoms. This could lead to the speculation that preictal administration may block or reduce the release of in ammatory peptides and the activation of related neurons, thus shorten the duration of headache and decrease the severity of headache.
Zolmitriptan is a selective agonist of 5-hydroxytryptamine 1B/1D receptor subtype [26], which is effective and safe for acute treatment of migraine and cluster headache. Nasal zolmitriptan is generally regarded as the acute therapy of CH with a level A recommendation by American Headache Society (AHS) guidelines, and oral zolmitriptan with a level B recommendation. The elimination half-life of zolmitriptan and its metabolites was about 3 hours [27]. The intervals of the drug administration of our patients were longer than the maintenance time and three half-lives of zolmitriptan, so the cumulative effect of drugs on the next treatment can be ruled out. As seen in previous CH studies, 57% of CH patients felt mild pain or painless after taking zolmitriptan 5mg for 30 minutes[28], while the rates in our study were 56-70% after taking zolmitriptan 2.5mg for 30 minutes. This may imply that preictal treatment of eCH could achieve the same or even better curative effect with a smaller dose zolmitriptan.
The possible reasons for the effective treatment of CH in the preictal phase were probably because that zolmitriptan was shown to be inhibit timely the release of vasoactive peptides via central and peripheral 5-HT1B receptors, and thus stop the cascade of neurogenic in ammation earlier. It regulated the intracellular calcium promptly so as to block the synaptic transmission between the terminals and body of peripheral trigeminal nerve neurons, as well as inhibited the pain information transmission mediated by CGRP, substance P (SP) and glutamate. Moreover, zolmitriptan played a therapeutic role by constricting the blood vessels and inhibiting the dilation of blood vessels. Besides, we could not ignore the presence of nausea. In this study, 10.4% of the patients reported nausea in the preictal stage and 60% in the ictal phase. The symptoms of nausea and vomiting in migraine patients led to delayed gastric emptying, which affected the absorption of oral drugs [29]. We speculated that the drug absorption in the preictal phase may be more complete than that in ictal phase.
There were some limitations in our study. Firstly, it was a single center observational study with a small sample size. The incidence rate of disease was lower than that in European and American countries and the research cycle was limited. In addition, due to ethical limitations, we could not compare the e cacy of prodromal and delayed administration. Further randomized, placebo-controlled studies are needed to assess the exact therapeutic effect in the preictal phase, and the results will be more convincing.

Conclusion
Our study demonstrated that preictal phase was frequent in eCH patients and preictal duration was positively correlated with the headache duration. After oral administration of zolmitriptan in the preictal phase, the severity of headache was lower and the duration of headache was shorter than that at baseline, and the therapeutic effect was consistent in a cluster period without any serious adverse reactions. Since the origin of CH attacks is unresolved, our study of preictal symptoms and early abortive treatment strategies could contribute to the understanding of CH pathophysiology. The locations of headache attack were multiple and the proportions were orbital 62.5%, temporal 60.4%, supraorbital 52.1%, frontal 43.8%, occipital 27.1%, parietal region 14.6% and neck 4.2%