After the first clinical studies [4, 5, 6] that showed the good prevalence of CAS during migraine attack, Obermann et al. [7] conducted an epidemiological study on the presence of unilateral CAS in the adult general population which found a prevalence of 25.2% of them (ratio M/F 1:2), more than 50% had at least two CAS, in which conjunctival injection and tearing were more represented. In addition, CAS were more frequently associated with unilateral headache and more severe pain attack, than those who didin’t have these symptoms. Very recently the Danish group’s study [28], conducted on population, reported prevalence of 57% of CAS (at least one CAS) and 31% with two or more CAS by a questionnaire while the prevalence reduced to 44% of one CAS and 22% with two or more CAS by semi-structured interview. The facial/forehead sweating was the most common symptom (39%) followed by lacrimation (24%). They reported weak correlation between symptoms and no clear clustering of them.

These data [7, 28] confirmes those coming from the clinical series [4, 5, 6, 8, 9, 10, 11, 12, 13, 14, 15, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27]. Further the prevalence may be sometimes undervalued because some researches only look for the unilateral CAS while in clinical series they may be also present bilaterally as well as migraine pain may have bilateral lateralization. In Table 1 the studies [7, 16, 28] on population are summarized.

Clinical studies show an obvious higher prevalence of CAS up to over 70% due to case selection, however in the clinical experience the CAS are often not reminded (maybe are mild), if not focused the attention, and that can probably explain the great variability of data and also the reported prevalence below 3.2% in Turkish database [16], as the same authors suggest. Danish researchers propose that future studies must include subjects with at least two CAS to define CAS migraine+. The prevalence of unilateral CAS varies between % in adults but several studies report that CAS occur bilaterally. The prevalence in women is confirmed and chronic migraine has often a higher prevalence of CAS than episodic migraine. In Table 2 [4, 6, 8, 11, 15, 17, 18] and Table 3 [17, 18, 19, 20, 22, 23, 24, 25, 26], the clinical series of adult migraineurs are summarized.

There are only 4 studies conducted in pediatric age, they don’t show a predilection for female sex (in prepuberal age male/female ratio is about equal in migraine), but the prevalence is in a range between 40% and 67% moreover they are present in about 50% of children under seven years, often bilaterally. Pediatric studies [12, 13, 14, 21] are summarized in Table 4 (Ref. [12, 13, 14, 22]).

The diagnostic criteria of the IHS classification [2], used in the TACs, identify the following signs/symptoms as an expression of the autonomic trigeminal reflex: lacrimation, eyelid edema, conjunctival injection, nasal congestion, rhinorrhea, face and forehead sweating, eyelid miosis and ptosis. These CAS are those usually searched in studies on migraines, however several authors have researched and proposed other symptoms in consideration of the parasympathetic/sympathetic innervation involved in the head, which are frequently present and often associated with others [12, 13, 14, 26]. Those reported are red ear, facial flushing, foreign body sensation in the eye, aura fullness, and most recently throat swelling and voice change. Fig. 1 illustrates the different signs/symptoms, reporting the range of prevalence of each single sign in the literature.

The prevalence of the single cranial symptom is variable in literature and probably depends on various factors. In general lacrimation and forehead/facial sweating are most represented, while myosis and ptosis are the least present (see Tables 1,2,3 and Fig. 1). It is probably due to easier recognition of the parasympathetic activation (lacrimation, conjunctival injection) compared to sympathetic hypoactivity (ptosis) by the migraineurs. We may affirm that often at least 50% of CAS+ migraine sufferers report more than one CAS in the course of an attack.

There is no particular difference between the pediatric and adult age, this data leads us to think that this phenotype is part of the pathophysiological aspect of these migraine sufferers and it does not appear related to the worsening of the disease in duration or severity of attacks because it is already present in a proportion of children under 7 years [14]. Certainly further prospective studies on a larger population are needed to clarify this important aspect.

Although different topographical areas of the head are involved (eyes, face, nose, ears, etc.) and functional areas of the autonomic control of the head (secretory, vasodilatory, sympathetic, etc.), the studies carried out fail to highlight a mode of association between them or a tendency to form specific clusters even if some authors favor these hypotheses [13, 20] while the most recent Danish study [28] does not support these conclusions.

In these studies that report the intensity of CAS in the course of a migraine attack, it is usually referred to as milder than that found in individuals with cluster headaches [8, 28], however this is a fact that must be confirmed because it may be due to several reasons, see the retrospective nature of the studies.

An interesting study by Viana et al. [37] on the variability of the clinical migraine phenotype recording three consecutive attacks, highlighted that, where CAS were present, they tended to occur in all three attacks in a percentage equal to 70% higher than those recorded for classic symptoms such as pain intensity, nausea, quality of pain, one-sidedness, photo-phonophobia, supporting the idea that in CAS+ migraineurs they are part of the specific cluster of migraine attack. The small number of the sample and the limited number of attacks recorded do not allow definitive conclusions. Lai et al. [8] showed that consistency of CAS in migrainous attacks was less than that in Cluster Headache. The study also showed that the CAS were milder and more bilateral in migraine subjects compared to the cluster headache. Their data were collected retrospectively.

Another study [27] evidenced that CAS+ migraineurs had significantly a greater central sensitization scored by the validated central sensitization inventory (CSI), suggesting a role of cranial parasympathetic activation to activate the central sensitization. This study support the Yarnitsky et al. [38] study that suggested that the role of cranial parasympathetic outflow to sensitize the intracranial nociceptors and to induce parasympathetically indipendent allodynia by sensitizing central trigeminal nociceptors.

Further indications on the correlation between CAS and pathophysiological mechanisms come from two studies that highlight how CAS are more present in migraine sufferers with attacks related to a seasonal cyclicality (also presenting greater disability) [15] and their correlation with cigarette smoking occurring more frequently in smoking migraineurs [9]. These two aspects, cyclicity of attacks and positive history of cigarette smoking, are part of the characteristics reported in subjects with cluster headaches and therefore suggest a link between the two groups and the involvement of similar anatomical-functional structures [9].

The study of Karsan et al. [39] suggests that the pain is not necessary to activate the CAS, as also showed for the Cluster Headache, supporting the hypothesis that CAS are not simply due to more severe migrainous attacks with a cascade involvement.

Almost all studies tried to correlate the presence of CAS in migraine attacks with the other clinical features of migraine. Several studies have shown a positive correlation between CAS and chronic migraine [18, 25], history of disease [6, 11], frequency of attacks [13, 21, 26], severe intensity of pain [4, 17], duration of attacks [6, 17], unilaterality of pain [4, 11, 16, 17, 25, 26], allodynia [17, 22], phono/photophobia [6, 17] and osmophobia [22]. However the different studies, compared among them, often conflict with each other, not confirming the correlations reported by the other studies and vice versa. The Danish group’s [28] study showed a weak correlation between symptoms, no more 0.41, and no clear clustering of symptoms. This is probably due to different selection of samples (prevalence of episodic versus chronic migraine), type of interview, age of the sample, symptom examined, etc. For this reason, although the literature on the subject is now fairly extensive, we agree with the conclusions by the Danish group [28] to study better this possible endophenotypic migraine. They suggest precise diagnostic criteria, differentiated according if they are aimed at genetic and/or epidemiological studies or clinical and/or pathophysiological studies.

Finally the involvement of cranial autonomic activity in migraineurs suggests that in this subgroup of migraineurs general autonomic activity may be more present than those with CAS. Raieli et al. [13] reported that pediatric migraineurs with cranial autonomic symptoms had more frequent general autonomic symptoms: vasomotor in 56% and GI dysfunction such as abdominal pain, cramping, early satiety, and persistent fullness in 42%. Recently Togha et al. [25] report that visceral symptoms were present about 52% of migraineurs that also reported CAS. However the authors do not precise the correlation between CAS and visceral autonomic symptoms.

The presence of a possible endophenotypic subgroup in migraines obviously led to the question if this could have any significance in response to treatment.

Barbanti et al. [5] and others [40, 41] suggested a greater response to triptans in CAS+ migraine subjects based on large-scale recruitment of peripheral neurovascular 5-HT1B/1D receptors consequent to the activation of the trigeminal autonomic reflex. Further they [10] supported their hypothesis with a double-blind study with Rizatriptan on CAS+ migraine sufferers demonstrating large gains at pain free at 2 hours.

Sarchielli’s study [35], made on migraine responders and non-responders to Rizatriptan, is interesting. The authors showed that migraine responders had more CAS than non-responders, as well as having more severe and one-sided pain. In addition, baseline vasoactive intestinal polypeptide (VIP) level was higher and decreased after treatment in CAS responders.

Further confirmation that triptans are more effective in migraine sufferers with CAS+ comes from the study by Viana et al. [29] using frovatriptan, resulting their presence as a predictor of response together with unilateral pain, the presence of phonophobia and one or more prodromes.

Curiously, but perhaps not too much, this better response to triptans in CAS+ migraine sufferers was not found in a study conducted in a pediatric population [30]. This data further confirms that the results from the studies in adults cannot be adapted tout court in the pediatric population.

Considering the proportion of subjects with chronic migraine CAS+, some authors [32, 33, 34] suggest to evaluate the better efficacy of Onanbotulinum toxin treatment on this population into account that some studies suggest different responses to this treatment based on the different clinical phenotype such as quality and direction of pain migraine [42].

Anedoctical case report showed a good efficacy of oxygen in a patient with severe migrainous attacks with CAS, further linking migraine and cluster headache [31].

Finally to our opinion it will be important to evaluate the efficacy of Antibodies anti-calcitonin gene-related peptide (CGRP ) in this subgroup of migraineurs, however remembering the importance of VIP levels in this population, while a study showed no difference between CGRP levels in migraineurs CAS+ and CAS– [43].

The prevalence of CAS in subjects with chronic migraine versus those with episodic migraine [8, 19, 20, 21, 22] suggested that the increased frequency of attacks may induce sensitization and greater involvement of the autonomic component due to the greater severity of disease. However, the evidences (1) that CAS, can also be induced in the absence of pain [39], (2) that this difference between episodic and chronic migraines is not always present in the studies, and (3) that they are already present in pediatric age and with no significant differences between preschool and school age children [12, 13, 14], also suggest the alternative hypothesis that CAS are an independent endophenotypic migraine subgroup. The absence of short- and long-term prospective follow-up studies on large samples do not allow for precise conclusions. However, we report an our long-term follow-up study [44] conducted on a small sample of migraine children under 6 years suggesting that CAS may be a predictor of the persistence and worsening of migraine syndrome over time and suggest early treatment as pharmacological as no-pharmacological strategies.

According to the 3rd IHS classification [2], migraine can be diagnosed on these main clinical aspects that underlie specific pathophysiological mechanisms: (1) temporal recurrence of the attacks; (2) duration of the painful attack (not less than 2 hours including the pediatric age); (3) the characteristics of pain in terms of unilateral lateralization, pulsating quality, pain severity and increasing with routine physical activity; (4) involvement of the general vegetative system (nausea and/or vomiting) and intolerance of sensory channels (photophobia and phonophobia). The migrainous patient, due to the current absence of specific biomarkers, is identified on the basis of the association of these clinical aspects which must not necessarily be present all in the single patient. The subject identified as migrainous will be included in the pharmacological and functional studies, however it is evident that, already on the basis of these diagnostic criteria, migrainous subjects with different clinical characteristics (for example duration of the attack, quality of pain, bilateral pain, vomiting present or absent etc) can be included in the studies, being able to arise conflicting results.

Indeed, migraine is a very complex disorder that involves multiple structures of the nervous system and has a complex clinical heterogeneity not sufficiently highlighted by the current diagnostic criteria. Over the years, the presence or absence of clinical features such as allodynia, osmophobia, general prodromes and autonomic symptoms can affect the therapeutic response and clinical outcome of the individual migraine sufferer.

In this general premise the attention placed, about 20 years ago, on the presence of CAS in migraine attacks (well described anecdotally in migraine patients) must still be placed. They are interpreted as a clinical expression of the hypothesized trigemino-autonomic reflex and they are the basis of the distinction of TACs, as an autonomous group of primary headaches of the IHS Classification. Their presence in a conspicuous proportion of migraine sufferers, both pediatric and adult, raises questions for the physiopathological aspects of migraine, on the relationship with TACs on the evidence of different behaviors in the therapeutic response and in abortive therapies of the migraine attack and on the evolution over time of the migraine syndrome in the individual patient. As for allodynia, they must be also identified and accurately described in the phenotypic characterization of the migraine attack in our patient, suggesting a more adequate intervention.

They, together with other clinical features of the migraine attack, such as allodynia, osmophobia, migraine aura, the quality of pain, general symptoms dued to para/orthosympathetic activity, etc., as well as highlighting the multifaceted and rich expression of the migraine attack in different subjects, invite the doctor to make a more individualized diagnosis of the patient because his response to therapy and to the modification of the clinical course also depends on this.