Keywords

11.1 Migraine Pain

Migraine is a heterogeneous disease with a spectrum of symptoms accompanying the headache pain. The classic migraine headache is characterized by a moderate to severe unilateral throbbing pain. Several other sensory symptoms add to the burden of the headache pain, the most common being nausea and light, sound, and smell hypersensitivity.

Migraineurs may experience symptoms hours to days before the headache attack, as well as symptoms that outlast the headache [1]. Some migraineurs can even predict migraine headaches based on preceding non-headache symptoms. Common prodromal symptoms are tiredness, concentration difficulties, neck stiffness, and increased sensory sensitivity [2, 3]. Migraineurs could report prodromal symptoms several days before the headache started, but analysis showed that the predictive value of these symptoms was rather low until the last 12–24 h before an attack [3].

Allodynia, defined as “pain due to a stimulus that does not normally provoke pain” [4], appears to be an important clinical correlate for altered pain processing in migraine. Allodynia may be assessed by questionnaire, by bedside assessment, or by more detailed experimental quantification as further discussed below. Simple clinical assessment of allodynia may include examination with cotton swab, pinprick, and thermal stimuli, i.e., normally non-painful stimulations [5]. When evaluated by questionnaire, about 50–70% of migraineurs report allodynia during headache, and allodynia is associated with frequency and severity of migraine [6,7,8,9]. Seo and Park [10] explored the clinical significance of allodynia compared with photo-, phono-, and osmophobia and found that both the prevalence of chronic migraine or medication overuse headache, disease duration, and headache intensity were increased in migraineurs with allodynia compared to migraineurs without allodynia, regardless of concomitant hypersensitivity to light, sound, or smell. Accordingly, the presence of allodynia is clearly associated with quality of life and increases the burden of the disease.

A case report showed that, during a migraine attack, allodynia started on the same side of the head as the headache and then spread to the other side of the head and finally to the arm with a progressive increase in magnitude [11]. The authors suggested that this represented activation of peripheral nociceptive neurons, followed by sensitization of second-order spinothalamic neurons and lastly third-order thalamocortical neurons [11]. In one study, at least one of heat, cold, or mechanical ipsilateral trigeminal allodynia was present in 79% of migraineurs 3–4 h into an attack [12]. Only five of those 33 subjects had ipsilateral trigeminal allodynia without contralateral or non-trigeminal allodynia, and two had contralateral but no ipsilateral allodynia [12], thus not providing any clear evidence of sequential activation of first- to second- to third-order trigeminal neurons.

Activation of nociceptive neurons innervating pial, arachnoid, and dural blood vessels and large cerebral arteries and sinuses, combined with a change in central pain modulation, is believed to give rise to the migraine headache [13]. Dysfunction of central nervous system structures involved in modulation of excitability and pain may activate and sensitize the trigeminovascular pain pathway [14,15,16]. However, the driving force behind this cycling activation is still unknown. Findings from functional imaging studies suggest involvement of hypothalamic, thalamic, and brainstem networks [17,18,19,20,21,22].

11.2 Experimental Pain

A strictly objective measure of pain perception is not available. Neurophysiological tests can measure nociception, the neural process of encoding noxious stimuli [4], but the degree of nociceptive activity does not necessarily comply with subjective pain perception. To quantify pain semi-objectively, a battery of psychophysical neurophysiological tests may be used, commonly defined as quantitative sensory tests (QST). A QST protocol is considered a useful method for psychophysical assessment of sensory detection and pain perception [23]. The protocol may include assessment of detection thresholds, pain thresholds, suprathreshold pain, and pain modulation. Modalities used may be pressure (deep mechanical), tactile (superficial mechanical brush, pins, or filaments), vibration (not used for pain), thermal (heat and cold), electrical (bypassing receptors), visual (light), auditory (sound), and chemical (for the nasal or oral mucosa). Pain thresholds are defined by the external stimulus, e.g., in degrees Celsius for thermal stimuli. An important limitation to the QST is that the tests require cooperation from the subject to define the moment a stimulus is detected, perceived as painful, or to rate the degree of pain experienced in suprathreshold pain experiments.

The thermal part of a QST protocol is particularly helpful in diagnosing small fiber neuropathy, mostly as hypoesthesia and hypoalgesia but also allodynia and hyperalgesia. A pain threshold below the normal range is interpreted as allodynia. Hyperalgesia, defined as “increased pain from a stimulus that normally provokes pain” [4], may be shown by increased pain intensity score to suprathreshold pain stimulation. Both allodynia and hyperalgesia may be due to peripheral sensitization, central sensitization, or both, although by separate and multiple mechanisms [24]. In example, failure of the central pain inhibition system to properly attenuate noxious stimulation may result in hyperalgesia, while failure to inhibit crosstalk between sensory modalities may result in allodynia [25]. More sophisticated QST measures may be applied to assess endogenous pain inhibitory function and endogenous pain facilitatory processes [26]. Conditioned pain modulation utilizes two concurrent noxious stimuli at separate body parts in a “pain inhibits pain” model to measure central pain inhibition. Temporal summation of pain uses repetitive nociceptive stimuli at a frequency of more than three per second to assess pain facilitation. Decreased conditioned pain modulation and increased temporal summation of pain indirectly indicates central sensitization. Thus, different QST findings might provide insights in the underlying pathophysiology.

11.3 Experimental Pain and Migraine

Several studies have investigated responses to experimental pain in migraine. Most of these studies compared responses from migraineurs in the interictal phase and controls, but some also compared responses between migraine phases, migraineurs with or without aura, or episodic and chronic migraine [27, 28].

A recent meta-analysis of QST and migraine identified 109 articles eligible for qualitative analyses [29]. Nahman-Averbuch et al. [29] provided a comprehensive overview of pressure, mechanical, heat, cold, and electrical detection and pain thresholds, as well as suprathreshold pain and pain modulation. The meta-analysis showed lower pressure and heat pain thresholds and higher suprathreshold cold pain ratings in migraineurs compared to controls. Another meta-analysis of pressure pain thresholds over the cranio-cervical region demonstrated comparative results, i.e., lower pressure pain thresholds in migraineurs compared to controls [30]. The studies included in both meta-analyses compared mainly migraineurs in the interictal phase with controls. However, when studies had measurements from multiple migraine phases, the data were collapsed and analyzed as a merged migraine group compared to controls. Thus, the meta-analyses compared migraineurs to controls irrespective of migraine phase, although the findings mainly are representative for the interictal phase. In general, migraineurs seem to be slightly more sensitive to painful stimuli between attacks compared to controls, although the effects are small [27, 28]. Also, some studies have shown increased temporal summation [31, 32] and less efficient conditioned pain modulation [33, 34] in migraineurs in the interictal phase compared to controls, suggesting central sensitization. No experimental pain test has so far proved reliable in distinguishing between persons with and without migraine. However, a multitude of factors may influence sensitivity in individual patients and contribute to the variation between studies. As discussed below, some of the variations may be explained by cyclical alterations related to proximity to the previous and next attack and some by migraine subtypes.

11.3.1 Pain Perception by Migraine Phase

Only a few longitudinal studies have examined experimental pain sensitivity in the preictal phase (Table 11.1). Neither pain intensity ratings by laser stimulation [35] nor pain scores to painful intranasal ammonia stimulation [19] seem to be altered in the preictal phase compared to the interictal phase. A study by Sand et al. [36] demonstrated decreased heat and cold pain thresholds in the preictal phase compared to the interictal phase. The effect was present when the preictal phase was defined with a 24-h limit, but not with a 72-h limit. Another study that analyzed heat, cold, and pressure pain thresholds with a 48-h preictal limit found no differences [37]. Apparently, different limits may be the source of the discrepant results. However, a follow-up study with a 24-h limit did not reproduce these findings as heat and cold pain thresholds did not change from the interictal to the preictal phase [38]. One explanation might be that preictal recordings were closer to the attack in the study by Sand et al. [36] than in the follow-up study [38], although the latter showed no association between pain thresholds and days to next attack. In contrast, Schwedt et al. [39] found a correlation between heat pain thresholds and time to next attack, as pain thresholds at both the arm and head decreased closer to the attack. Also, pain intensity ratings to suprathreshold heat stimulation have been shown to gradually increase during the interictal period toward the next attack [38]. Thus, studies indicate gradually increasing pain sensitivity in the interictal phase toward the next attack with a more pronounced hyperalgesia during headache.

Table 11.1 Pain perception by migraine phase
Full size table

A study of heat pain intensity scores found an interictal correlation between pain scores and time to next migraine attack and a distinct increase during headache [38]. However, in the 24 h preceding the attack, a subtle decrease of pain scores was present, interpreted as preictal hypoalgesia. These results suggest that significant central events affect processing of pain on the day before headache. Hypothalamic activation has been shown in the preictal phase [17, 18]. Depending on the receptor activated, regions in the hypothalamus may provide either pro- or antinociceptive effects on trigeminal nociception [40]. Thus, it is plausible that preictal hypothalamic activation may cause a transient hypoalgesic effect by increased descending pain modulation. The antinociceptive effect seems to have an effect mainly on suprathreshold pain scores as pain thresholds have been shown to increase or remain unaltered in the preictal phase [36, 38].

Alterations of pain perception are more pronounced in the ictal phase. Studies have shown reduced pain thresholds to either heat, cold, mechanical, visual, and auditory stimulation, increased pain scores to tonic heat, and decreased pain thresholds tested by laser stimulation during attack compared to between attacks (Table 11.1) [12, 38, 41,42,43,44]. There are some contradicting findings, but the overall impression is an increased pain sensitivity during headache compared to the interictal phase, which corresponds well with the increase in allodynia and other sensory symptoms during the ictal phase.

Studies of the postictal phase have not shown any differences compared to the interictal phase [35, 37, 38]. When compared to the ictal phase, postictal normalization of cold pain thresholds has been shown [38], indicating a rather fast restoration of pain perception back to interictal levels.

11.3.2 Pain Perception by Migraine Subtypes

Studies comparing experimental pain in migraineurs in the interictal phase and controls have shown variable results, either hypersensitivity or no differences, but never hyposensitivity [27, 28]. Some subgroups may be more hypersensitive than others; for instance, migraineurs with non-sleep-related migraine attacks had lower thermal thresholds than controls [45], while less slow-wave sleep was associated with higher pressure pain thresholds [37].

Subjects with chronic migraine (more days with than without headache) seem to have more allodynia and lower pain thresholds compared to episodic migraineurs, indicating a relationship between altered pain perception and headache frequency [9, 31, 46,47,48,49]. However, other studies have neither shown any differences in mechanical or thermal pain thresholds between chronic and episodic migraine [50] nor a relationship between pressure and thermal pain thresholds and migraine frequency [39, 51]. Disease severity may also be of importance, as headache history duration has been shown to modulate cold pain thresholds [36]. As suggested by Peng and May [27], the increased pain sensitivity in chronic migraine may be due to a higher probability of being tested close to the ictal state compared to episodic migraine with longer interictal periods. Other important factors that may increase pain perception in chronic migraine seem to be increased headache severity and level of drug intake [28].

A twin survey suggested that migraine with and without aura are distinct disorders [52]. On the other hand, the International Classification of Headache Disorders, third edition, states that the same person may have both diagnoses [53]. Nevertheless, the few studies that have compared pain thresholds between migraineurs with and without aura have not shown any differences for thermal [36, 38], electrical [32], light [43], or sound pain thresholds [44]. Russo et al. [54] compared heat pain intensity in groups divided by migraineurs without aura and without ictal allodynia, without aura but with ictal allodynia, with aura but without allodynia, and controls. The study showed no differences between the migraine subgroups or compared to controls and no association between pain intensity and migraine severity. Granovsky et al. [55] compared migraineurs with and without aura and found increased temporal summation of mechanical pain stimulation in migraineurs with aura, but no difference in heat and mechanical pain thresholds, or conditioned pain modulation. Perenboom et al. [56] quantified visual allodynia and demonstrated higher scores in migraine with aura compared to without aura and in chronic compared to episodic migraine. Thus, visual stimulation may be better suited to differentiate between migraine with and without aura compared to thermal and pressure pain, although Vanagaite et al. [43] did not find altered visual pain sensitivity in migraine with aura compared to without aura.

Studies of pain thresholds in children with migraine are scarce, but resemble findings shown in adults [28]. Some studies have shown differences in mechanical pain thresholds [57], pressure pain thresholds [58], and laser-evoked pain thresholds [59] compared to controls, but conflicting results exist [60, 61]. A recent study demonstrated increased heat pain intensity scores in adolescents with migraine compared to controls, but no difference in conditioned pain modulation [62].

11.4 Conclusion

Pain perception alternates within the migraine cycle (Fig. 11.1). Thresholds gradually decrease toward the next attack with a distinct reduction during headache. What happens during the hours to days before the headache starts is still poorly understood. A few longitudinal studies have shown both decreased pain thresholds and paradoxical decreased pain intensity ratings in the preictal phase. The symptoms associated with the preictal and ictal phase, in addition to recent functional imaging findings, may suggest thalamocortical alterations by hypothalamic modulation as a generator of the observed preictal hypoalgesia. Findings of cyclical alterations of pain perception support the theory that migraine is a cyclic disorder of the central nervous system related to global alterations of brain excitability and homeostasis.

Fig. 11.1
figure 1

Threshold changes during a migraine cycle. The changes in sensory thresholds over time and their correlation with the clinical symptoms. The undulating threshold in healthy controls reflects the high day-to-day variance as reported in the literature [63]. In this figure, only the phasic changes among migraineurs are depicted; however, the day-to-day variance also stands true to the migraineurs. (1) Definition in the International Classification of Headache Disorders, third edition. (2) No consensus: Certain studies showed lower threshold among migraineurs than healthy controls; others showed no difference. (3) Two studies showed preictal threshold lower than interictal threshold among migraineurs. (4) Hypothetical: No study examined the sensory threshold in the postictal period in comparison with the ictal period (From Peng KP, May A. Migraine understood as a sensory threshold disease. Pain. 2019;160(7):1494–501. doi:https://doi.org/10.1097/j.pain.0000000000001531. Reprinted with permission)

Full size image