Variability of the characteristics of a migraine attack within patients

Abstract

Background

Migraine attacks may present different features in different patients and also within the same patient. The percentage of patients reporting stereotyped attacks and those reporting attacks with different phenotypes has not been the object of specific investigations.

Objective

The objective of this article is to evaluate the percentage of migraine patients reporting the same characteristics, in terms of phenotype and response to symptomatic medications on three consecutive migraine attacks.

Methods

Thirty patients with migraine without aura prospectively recorded the features of three consecutive attacks in a headache diary. Characteristics recorded were: pain intensity, presence of nausea, vomiting, photophobia, phonophophia, osmophobia, allodynia, cranial autonomic symptoms (at least one), and premonitory symptoms. Patients were allowed to take frovatriptan as symptomatic medication, whose efficacy was evaluated as the two hours pain-free status.

Results

None of the patients presented identical characteristics on the three studied attacks. This was still the case if we reduced the number of variables evaluated from 11 to seven of the eight core features indicated by the ICHD. Considering just six variables: unilaterality and quality of pain, presence/absence of nausea, vomiting, photophobia and phonophobia, only two patients (6%) had identical features on three consecutive attacks.

With respect to the response to frovatriptan, 39% of patients had the same response, either positive (i.e. pain free after two hours) or negative (i.e. not pain free after two hours) on three consecutive attacks.

Conclusion

Migraine attacks show a high variability not just among patients, but also within the same patient. Our data indicate that stereotypy of attacks is uncommon, and reinforces the underlying logic of the current operational classification system.

Introduction

Migraine is the third most common disorder in the world and the seventh in term of disability (1). It is characterized by episodes of moderate to severe headache that is mostly unilateral, throbbing and aggravated by routine physical activity. Headache is accompanied by non-headache symptoms, such as nausea, vomiting, photophobia, and phonophobia (2). In addition, osmophobia (3,4), cranial autonomic symptoms, such as conjunctival injection, lacrimation, nasal congestion, eyelid edema (5) and allodynia (6–8) are all relatively common in migraine. Migraine attacks can also be preceded by premonitory symptoms, such as tiredness, difficulty in concentrating, stiff neck, photophobia, phonophobia, irritability, and difficulty with thoughts (9–11). It is common experience that all these features of migraine attack can be very different from patient to patient. Yet, little information is available on the variation of the phenotypic presentation of migraine attacks in the same patient over time.

With respect to the response to antimigraine medication, it can vary among patients (12) but also in the same patient. This latter aspect has been shown in studies in which the response to medication has been tested on multiple attacks, as summarized in the review by Ferrari and colleagues on 53 randomized controlled trials on triptans (13).

The aim of this study is to evaluate the percentage of migraine patients reporting attacks with the same characteristics, in terms of phenotype and response to a given medication (frovatriptan 2.5 mg p.o.) on three consecutive attacks. The data have been presented in a preliminary form to the 4th European Headache and Migraine Trust International Congress (14).

Methods

We recruited 39 consecutive patients affected by migraine without aura at the Headache Science Center of “C. Mondino” National Neurological Institute of Pavia, Italy. The study received the approval by the local Ethics Committee and all patients signed an informed consent form. Enrollment started in October 2012 and ended in March 2014. Inclusion criteria were: i) headache fulfilling International Classification of Headache Disorders, second edition (ICHD-II) criteria for migraine without aura from at least one year with a frequency of attack among one and five per month and ii) age between 18 and 65 years. Exclusion criteria were: i) previous or current diagnosis of medication-overuse headache, ii) current diagnosis of migraine with aura or chronic migraine, iii) contraindications to be prescribed with triptans, iv) previous use of any triptan, v) presence of other types of headache attacks (i.e. tension-type headache) if the patient is not able to distinguish them from migraine without aura attacks, vi) use of opioids, vii) use of a migraine-preventive medication, viii) patients suffering exclusively from migraine attacks with onset at wakening, ix) history of psychiatric disorders.

At the first visit, after the diagnosis of migraine without aura, patients were prescribed frovatriptan 2.5 mg p.o. as a symptomatic medication and were given an ad hoc headache diary in which to record prospectively the features of three consecutive migraine attacks. Patients were asked to answer the questions at the moment of symptomatic medication intake, which they were advised to take as soon as they recognized the headache as a migraine. Characteristics of the attack that were recorded were: unilateral location, pulsating quality, pain intensity (mild, moderate, severe), presence of nausea (N), vomiting (V), photophobia (PT), phonophobia (PN), osmophobia (O), allodynia (A), presence of cranial autonomic symptoms (CAS, namely conjunctival injection and/or lacrimation, nasal congestion and/or rhinorrhea, eyelid edema, forehead and facial sweating, miosis and/or ptosis), presence of premonitory symptoms in the previous 24 hours (namely yawning, vertigo, tiredness, mood changes, neck stiffness, N, PT, PN, O, food craving, thirst, difficulty concentrating, reading, or speaking); and time of the triptan intake from the onset of pain. Allodynia was assessed by receiving a positive answer to the question “Does a non-painful stimulus (such as touching or washing or shaving or brushing or wearing glasses/contact lenses/earrings) produce pain on your face/head?” Patients were asked to record the pain intensity two hours after the triptan intake, and to report in the diary the whole duration of the attack, i.e. when all migraine symptoms stopped. Triptan efficacy was assessed as the two-hour pain-free endpoint. If the patients did not reach this endpoint, they were allowed to take an analgesic as a rescue medication. Upon completion of the task (recording and treatment of three consecutive migraine attacks) or, at the latest, three months after the first visit, patients returned to the headache center and their diaries were discussed with the neurologist. In order to evaluate the variability of the characteristics of a migraine attack within patients, we assessed i) the percentage of patients in which a given characteristic recurred with the same modality—for all the features the modalities were yes/no, except for intensity (mild, moderate, severe), CAS (at least one) and PS (at least one); ii) the percentage of patients in which a given number of migraine attack features (from two to 11) recurred together with the same modalities (i.e. in all the three attacks pain intensity was moderate, pulsating pain was present and nausea was absent).

Results

Thirty-nine patients were enrolled. Nine dropped out either because the number of attacks recorded was less than three or patients’ diary compilation was insufficient for evaluation.

Thirty patients recorded successfully in the diary the features of three consecutive migraine attacks treated with frovatriptan 2.5 mg p.o. for a total number of 90 attacks. These patients had the following characteristics: 83% were female (25/30), mean age 32.9 ± 8.3 years (range: 19–47), mean age at onset of migraine 21 ± 11 (range 5–47), mean duration of illness 16.0 ± 9.5 years (range 1–30), mean frequency of migraine without aura attacks was 3.4 ± 1.1 per month (range 1–5). In the past, they had treated their migraine attacks with nonsteroidal anti-inflammatory drugs (NSAIDs) or paracetamol. Out of 30 patients, 28 suffered exclusively from migraine with aura attacks. Two patients also suffered from infrequent episodes of tension-type headache and experienced one tension-type attack each during the observation period after enrollment. The patients clearly differentiated these two attacks of tension-type headache from migraine and therefore did not take acute antimigraine medications nor recorded the attacks in the ad hoc diary. In the study period, they also had three migraine attacks each, which were treated with frovatriptan and were recorded in the ad hoc diary.

Migraine attack features

The features of the 90 attacks were represented with the following frequency: pain was unilateral in 57% of the attacks (n = 52), quality was pulsating in 40% (n = 36), intensity was mild in 34% (n = 31), moderate in 48% (n = 44), and severe in 16% (n = 15). Nausea was reported in 33% of attacks (n = 30), vomiting in 2% (n = 2), photophobia in 52% (n = 47), phonophobia in 41% (n = 37), osmophobia in 17% (n = 16), allodynia in 20% (n = 18), at least one CAS in 21% (n = 19), and at least one premonitory symptom in the previous 24 hours in 63% (n = 57). The frequency of each single premonitory symptom varied from 2% (two of 90) for craving/thirsty to 35% (32/90) for tiredness. All the attacks lasted less than 72 hours. No patients presented any migraine aura associated with migraine attacks.

Variability of migraine attack phenotype

Each feature recurred with the same modality during the three attacks with the following frequency: unilateral pain in 43% of patients (13/30), pulsating quality in 57% (17/30), intensity in 23% (7/30), nausea in 60% (18/30), vomiting in 93% (28/30), photophobia in 60% (18/30), phonophobia in 57% (17/30), osmophobia in 77% (23/30), allodynia in 77% (23/30), CAS in 70% (21/30), and premonitory symptoms in 43% (13/30) (Figure 1).

Figure 1.

Frequency with which a characteristic occurs with the same modality over three attacks in 30 patients. P: pain; N: nausea, vomiting; PT: photophobia; PN: phonophobia; O: osmophobia; ALLO: allodynia; CAS: cranial autonomic symptoms; PS: premonitory symptoms. For all the features the modalities were yes/no, except for intensity (mild, moderate, severe), CAS (at least one) and PS (at least one).

Out of 30 patients, no patient had identical characteristics for the 11 variables on the three studied attacks (Table 1). The results were unchanged even when we took into consideration only seven variables: the three characteristics of head pain (unilateral pain, pain intensity and throbbing quality) and the four associated symptoms described in D criterion of the ICHD, third edition beta (ICHD-III beta) (presence/absence of N, V, PT, PN). When we considered only six variables (unilateral pain, throbbing pain, presence/absence of N, V, PT, PN), we had only two patients (6%) with three identical attacks.

Table 1.

Migraine attack features and intrapatient variability.

Unilateral location	Pulsating quality	Pain intensity (four-point scale)	Nausea	Vomiting	Photophobia	Phonophobia	Osmophobia	Allodynia	CAS	PS	Number (%) of patients with three identical attacks with respect to some of the features recorded at the moment of symptomatic medication intake
x	x	x	x	x	x	x	x	x	x	x	0 (0%)
x	x	x	x	x	x	x					0 (0%)
x	x		x	x	x	x					2 (6%)
			x	x	x	x					9 (30%)
x	x										10 (33%)

CAS: cranial autonomic symptoms (at least one); PS: premonitory symptoms (at least one).

Finally, when we considered just the four associated symptoms mentioned in ICHD-III beta (N, V, PT, PN), nine out of 30 patients (30%) had three identical attacks. When we considered instead only the two characteristics of head pain (unilaterality and quality), 10 out of 30 patients (33%) had three identical attacks. Out of the 13 patients who presented with at least one premonitory symptom in one attack, nobody had the same type of premonitory symptom(s) over three different attacks.

Response to treatment

Patients took frovatriptan 2.5 mg orally at a mean of 63 minutes after pain onset, when the median pain intensity was 2 (interquartile range: 1–3). In 30 out of the total 90 attacks (33%), the patient became pain free in two hours. If we consider the consistency of response to frovatriptan, just 39% of patients had the same response, either positive (i.e. pain free after two hours) or negative (i.e. not pain free after two hours) during three consecutive attacks.

Discussion

This prospective, diary-based study of migraine without aura features demonstrates that attacks are very different not just among patients, but also within the same patient. Out of 30 patients, no one had identical characteristics during the three studied attacks, even if we reduce the number of variables evaluated from 11 to seven out of the eight features indicated in criteria C and D for migraine without aura both of ICHD-II and ICHD-III-beta.

Considering just six variables: presence/absence of unilateral pain, throbbing pain, nausea, vomiting, photophobia and phonophobia, only two patients (6%) had identical features in three consecutive attacks.

Evaluating each single migraine feature, their consistency over the three attacks varied from 23% to 93%. We have to consider that the minimum, 23%, was related to the intensity of pain that has three degrees, while the maximum, 93%, was related to vomiting, which occurred just in two attacks (of two different patients) out of 90. Excluding these extremities, the range of consistency for an “average” bimodal feature varied from 43% to 77%. From this, we could infer that approximately six patients out of ten has consistency of presentation of a given migraine feature over three consecutive attacks.

To the best of our knowledge, this is the first study that systematically assessed the percentage of patients reporting migraine attacks with identical features in a given period. In a previous study, Nachit-Ouinekh and colleagues have shown that some characteristics of migraine attacks, those nine defined in the International Headache Society (IHS) classification, are relatively stable over time (15). In their study, the authors recorded retrospectively the features of a typical migraine attack in two discrete periods separated by a 12-month interval, although they did not evaluate phenotype variability over any single period. In another study, intrapatient variability of only one feature of the attack in a given period was considered (16). The study was prospective and evaluated the characteristic “intensity of pain” over several attacks, potentially up to 10, in the same patients. The study showed that the severity of the second attack was identical to the first, using a 0–3 scale, in 59.1% (n = 543) of patients who had experienced a mild first attack, 40.8% of patients who had experienced a moderate first attack, and 43.3% of patients who had experienced a severe first attack. We calculated the consistency of pain intensity across attacks in our sample and our results, though obtained on a smaller number of attacks, showed similar values. In our sample the intensity of the second attack was identical to the previous one in 50% of patients who had experienced a mild first attack, 46% of patients who had experienced a moderate first attack, and 50% of patients who had experienced a severe first attack.

Our results show also that attacks can be different in the same patient not just in terms of clinical features but also in response to a given symptomatic medication. This had already been demonstrated in other studies (13). Here we have provided some data that could help to understand, at least in part, such intrapatient variability in response to a given medication. Indeed we demonstrate that the same patient might have different attack phenotypes. It is reasonable to hypothesize that these different presentations are the expression of a differential activation of components of the pathophysiological migraine machinery. This differential activation may present dissimilar interactions with pharmacodynamic mechanisms of a given medication.

One limit of this study, with respect to the migraine features, is the possibility that some migraine attacks may have been stopped or relieved by the triptan before the complete phenotype could be expressed. This may have caused a reduction of the frequency of some features, i.e. nausea or photophobia, which is indeed lower than those reported in the literature (6). However, it seemed unethical to ask the patients not to treat attacks when the chances of efficacy are higher. This would also affect the response to medication. We do not know whether an evaluation of the attacks at the time of maximum symptoms expression would have changed our results in terms of phenotypic intrapatient variability.

Another potential limitation of our study regards the fact that our ad hoc diary did not foresee the registration of ICHD-II criterion 4: “pain aggravation by or causing avoidance of routine physical activity.” The inclusion of this additional variable would probably have increased the degree of heterogeneity of the attacks, but we felt that satisfaction of criterion 4 was not the priority in a population instructed to take the acute medication as soon as they recognized the headache as a migraine, in line with the latest edition of the Guidelines for Clinical Trials (17). To confirm the migraine nature of the attacks considered for the analysis, we revised the characteristics of the attacks recorded in the diaries and none of them fulfilled criteria for tension-type headache, nor for other types of primary headaches.

Another methodological issue of this study is that our findings cannot be directly extrapolated to the general migraine population, as patients were recruited from tertiary referral centers and were selected according to specific inclusion/exclusion criteria. However, we felt that strict procedures were necessary to limit confounding effects, and these procedures (i.e. diagnosis of migraine without aura made by a headache expert; need to fill out prospectively an ad hoc diary and the triptan naivety) could not be reliably applied in a population study. In this frame, the selection of migraine patients who have never used triptan was prompted by the observation that previous use of triptans could modify the response to a new triptan (12,18). One might think that triptan-naïve migraine patients are not representative of the general migraine population. This is not the case, at least for the Italian population, as shown in a relatively recent Italian survey on patients consecutively attending for the first time 10 Italian headache centers over a three-month period that reported that only 17.2% of migraine patients used triptans (19). Therefore our population reflects the triptan-naïve characteristic that is present in 82.2% of the population attending a headache center. It is noteworthy that the problem of migraine undertreatment does not involve only the Italian migraine population (20,21). With respect to demographic data, in our sample the age at onset of migraine is in line with that reported in the literature while our population has a lower male/female ratio (1:6 versus 1:3 in the literature) (22). Indeed the number of patients is too low for performing statistical analysis to evaluate if our sample is representative of the general population, yet this was not the main focus of our paper. Our focus was instead to study in detail the possible existence of an intravariability of migraine attacks presentation in a given period.

In conclusion, for the first time we provided prospective, diary-based data to demonstrate that migraine attacks are very different not just among patients, but also in the same patient. Deep phenotyping of their nature, together with molecular techniques and therapeutic outcomes (23), may provide both a better understanding of migraine and avenues through which we can treat patients more effectively.

Clinical implications

Here we demonstrated with a prospective, diary-based study that migraine attacks are very different not just among patients, but also in the same patient.

Deep phenotyping of attacks is a nodal step to further advance our understanding of migraine pathophysiology and to improve treatment efficacy.