Abstract
Sneddon's syndrome refers to the enigmatic association of ischaemic stroke and livedo reticularis. We review the Sneddon's syndrome literature examining the association of this condition with headache, including migraine. Case reports and series are stratified into two groups based on headache reference. In the group without a reference to headache, there are 208 persons, with a female to male ratio of 3 : 1. In the headache reference group, there are 175 persons, with a female to male ratio of 3.5 : 1. The proportion with headache in this second group is 58% (102 individuals), with headache described as migraine in 28 (27.5%) of the headache subjects, including six with migraine with aura. The frequency of headache is not significantly higher in persons with positive anti-phospholipid antibodies compared with the negative cohort (43% vs. 32%, P = 0.07). A review of the histopathological, radiological and serological data in Sneddon's syndrome and migraine underscores the plausibility of an association. Considered in the context of increased risk of stroke with migraine, a higher frequency of livedo in migraineurs with stroke, and the association of migraine and livedo reticularis, the question of whether livedo reticularis may be a risk marker for stroke in migraineurs is an area for further study.
Introduction
The association of ischaemic cerebrovascular disease and ischaemic dermatopathy was first described in 1960 by Champion and Rook (1). This came to be known as Sneddon's syndrome after the series reported by the British dermatologist, I. B. Sneddon, in 1965 (2). He described five women and one man who, between the ages of 20 and 42 years, presented with ‘cerebrovascular incidents which have been of limited and benign nature’ associated with livedo reticularis (LR). The reddish blue mottling of the skin in a ‘fishnet’ reticular pattern often preceded the cerebral manifestations by many years. Sneddon was unable to identify an underlying cause for stroke, and his eponymous syndrome usually denotes the pairing of LR in the absence of connective tissue disease, neoplasm, infection or an inflammatory condition. In recent years it has been reported that LR is common in persons being treated for migraine, and that migraineurs with LR more frequently report a history of stroke than those without LR (3). In another study from a dermatology clinic population, LR was found, among women, to be associated with migraine (4). Given that migraine has been widely reported to be a risk factor for stroke, especially in young women (5–13), the question arises as to whether Sneddon's syndrome is associated with migraine. Although it has been reported that headache is common in Sneddon's syndrome (14), the relationship of Sneddon's syndrome and migraine, and the meaning of such an association in pathogenic terms, has not been explored.
Methods
To determine the prevalence of headache and migraine in persons with Sneddon's syndrome we reviewed the literature using MEDLINE online database from 1966 through February 2005 and Sciences Citation Index (SCI) database from 1956 through February 2005, with key words (‘Sneddon's syndrome’, ‘livedo reticularis’, ‘migraine’, ‘headache’, ‘stroke’, ‘cerebrovascular disease’) for case reports and series detailing the clinical specifics of persons receiving this diagnosis. The ‘Related Articles’ feature of PubMed and the ‘Cited Reference Search’ in the Web of Science were also searched. Additionally, we checked the references of the selected articles to detect other eligible publications. We included all studies published in English, or with an English abstract, regardless of methodological characteristics and subjective quality criteria. When subjects were cited in more than one article, we included only the larger study in our estimations, in order to avoid duplication. Due to a paucity of clinical detail in many studies, case series were stratified into two groups based on whether or not headache was mentioned for any of the reported subjects, even if the headache history was negative. In each of the two groups, we calculated the gender ratio. In the headache reference group, we determined the proportion with headache, and with migraine, even if not further defined. We interpreted the lack of mention of headache for an individual as an absence of headache. Furthermore, we estimated the frequency of headache in individuals with anti-phospholipid (aPL) antibodies compared to those without. Every article was reviewed for histopathological, radiological and serological findings.
Results
We identified 97 case series and reports with 426 individuals with Sneddon's syndrome. In 42 case reports and 23 case series, containing a total of 208 individuals, the presence or absence of headache is not mentioned (15–64). The female to male ratio for this group is 3 : 1. In 19 of the case series, and 13 of the case reports, headache was referenced and this group contained 236 persons, with 61 duplicates, for a total of 175 persons. Table 1 lists the cases and series referencing a headache history for at least one individual in the series (2, 65–95). The ratio of females to males is 3.5 : 1 in the headache reference group. In this same group, the proportion of persons with headache was 58%, affecting 102 of 175 persons. We are unable to calculate the gender ratio for those with headache due to lack of clinical detail in the literature. Of the 102 persons with headache, 28 (28%) were described as having migraine, although only one case series (94) indicated the use of International Headache Society (IHS) criteria (96). The frequency of headache in 116 individuals with aPL antibodies was 50 (43.1%) compared with 31 (32.6%) in 95 individuals without aPL antibodies (P = 0.07).
Headache history in Sneddon's syndrome
M, Male; F, female; HA, headache; LR, livedo reticularis.
Range is specified if more than two cases in the study.
Cases previously reported.
Mean age of the cases.
Discussion
The finding of a high prevalence of headache, including migraine, in Sneddon's syndrome is intriguing, particularly in the context of studies linking LR to migraine (4), migraine to stroke (5–13), and demonstrating that among migraineurs, a history of stroke is more common in those with LR (3). There are, however, limitations to a retrospective review such as this, the most significant being the paucity of clinical data in the Sneddon's syndrome literature. We cannot ascertain whether there was systematic assessment of headache history in any of the case reports or series, and this review may actually underestimate headache frequency. We have calculated, based on approximately half of the reported cases (i.e. the headache reference group), that 58% of persons with Sneddon's syndrome had headache. This is similar to other estimates in the literature (14). If, however, all 383 subjects with Sneddon's were included in the headache frequency calculation, the proportion with headache would decrease from 58% to 27%. Another methodological limitation is that migraine, when reported, was usually not defined, with many of the reports before the establishment of IHS criteria. When the non-specific term ‘headache’ was used, we cannot determine what proportion of these may have been migraine. If all the headaches in the headache reference group (58%) were migraine, this would be greater than the prevalence of migraine in the general population (97), but there are no data to support this assumption. In the literature, 28 of the 102 headaches in the headache reference group (16%) were described as ‘migraine’, which is similar to female prevalence of migraine in the general population.
Within a headache clinic, analysis has shown an odds ratio (OR) of 4.4 [95% confidence interval (CI) 1.4, 13.9] for stroke within the cohort with LR (3). In a dermatology clinic, the association of migraine and LR was established in women with an OR of 2.3 (95% CI 1.1, 4.7) (4). It has been observed that in Sneddon's syndrome vascular headache frequently precedes the occurrence of LR by several years (72), and LR often precedes stroke by many years (2, 68, 72, 75, 82, 91). This temporal dissociation, coupled with the likelihood that Sneddon's syndrome is underecognized (2), may account for the fact that the relationship of migraine and Sneddon's syndrome has not previously been widely explored. Given that the Sneddon's literature does not allow one to draw conclusions on the comorbidity of migraine and Sneddon's syndrome, we have examined the literature documenting the histopathological, serological and radiological data from persons with Sneddon's syndrome to evaluate whether there may be common biological mechanisms to support the hypothesis of an association between these conditions.
LR results from focal impairment of blood flow of the small and medium-sized vessels at the dermis–subcutis border. The underlying cause of this dermatopathy in persons with Sneddon's syndrome has generally been attributed to non-inflammatory thrombotic vasculopathy rather than to inflammatory vasculitis (17, 46, 79). Although in some cases skin biopsies have been negative or non-specific (2, 17, 32), many reports describe endarteritis obliterans from intimal endothelial proliferation (1, 35, 48) and proliferation of medial smooth muscle cells (66). Digital artery biopsies in persons with negative skin biopsies reveal intimal hyperplasia, focal zones of absence or reduplication of internal elastic lamina and adventitial fibrosis (17). Inflammatory features are infrequently observed, and are usually not prominent (72, 73, 98). Some series have suggested a distinct histopathological time course, with early inflammatory reactions evolving into subendothelial proliferation, then to a late fibrotic stage (72, 73).
Brain histopathological reports in persons with Sneddon's syndrome are limited in number but suggest that a similar non-inflammatory endothelial process may underlie vascular lesions in the brain. Our review uncovered three autopsies (47, 50, 99) and five cerebral biopsies (24, 57, 79, 80, 89). Three of the biopsies contained foci of necrotic brain (24, 57, 89), but abnormalities of the leptomeningeal or superficial arteries were absent. Two biopsies demonstrated thrombosis and recanalization in meningeal and/or cerebral vessels without evidence of vasculitis (79, 89). There was only one report of a leptomeningeal biopsy showing granulomatous infiltration, suggestive of an inflammatory process (80). None of the three autopsies found evidence of inflammation. One showed diffuse cerebral vasculopathy, with musculoelastic hyperplasia and scattered fibrotic occlusion and recanalization (99). Another autopsy case showed leukoencephalopathy, multiple cortical infarcts and fibrosis of the intima of the vessels comprising the circle of Willis (50). A third autopsy reports multiple small cortical and subcortical infarcts with fibrotic occlusion and recanalization of medium-sized vessels in the leptomeninges and superficial white matter (47). Magnetic resonance imaging (MRI) data, with the most characteristic finding being small to moderate-sized infarcts located in the cortical and subcortical areas (40, 73, 92), support the histopathological finding of involvement of small and medium-sized vessels in the brain.
Vascular thrombosis and recanalization within skin and brain tissue suggest that coagulopathy may be an important contributor to the pathology of Sneddon's syndrome. Many persons with intraluminal thrombi have been found to have elevated aPL antibodies (98, 100), which are reported to be increased in one-third to one-half of those evaluated with Sneddon's syndrome. Based on this finding, Sneddon's syndrome has been characterized as an aPL syndrome, i.e. an autoimmune prothrombotic condition characterized by venous or arterial thrombosis. We found that in the Sneddon's literature the frequency of headache was not significantly higher in the aPL antibody-positive subset. Platelet activation has been documented in Sneddon's syndrome (16, 53) and the finding of increased antigen levels of von Willebrand's factor (101), a marker of endothelial dysfunction causing platelet activation, suggests a link between the vasculopathy and coagulopathy in this disorder. A number of other coagulation abnormalities have been reported, including elevated coagulation factor VII (61) and VIII activity (102), and decreased protein S (61, 92) and activated protein C resistance (23).
We are not aware of any pathological studies on skin in migraineurs per se, but in CADASIL (cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopthy), a condition associated with migraine and small strokes, laser Doppler flowmetry studies show an impairment of skin microcirculation related to endothelial dysfunction (103). There are only a few reports of brain pathological examination in migraine (104, 105). Despite the paucity of histological data in migraineurs, radiological data are consistent with small-vessel vasculopathy as a mechanism of migraine-related stroke. Several series report increased frequency of small white matter abnormalities on T2-weighted MRI in migraineurs compared with migraine-free cohorts (106–108). Hypercoagulability in migraine is controversial (109). The largest controlled study of aPL antibodies (specifically, anticardiolipin antibodies) in migraine did not confirm an association of aPL and migraine, although silent strokes were more frequent in the aPL antibody-positive migraine subset (110). Von Willebrand factor is elevated in migraineurs, particularly in those with a history of aura or stroke (111), suggesting endothelial dysfunction as a mechanism of primary haemostasis and non-inflammatory endarteritis obliterans.
This review underscores reports that found headache is frequent in Sneddon's syndrome (72, 75, 84, 94). With regard to the association of migraine and Sneddon's syndrome, epidemiological data are inconclusive. Biological data, however, support the plausibility of a migraine–LR–stroke association. Although the findings of our review are provocative, rather than definitive, it may prompt thorough investigation of the possible link between migraine and Sneddon's syndrome, with an eye on establishing whether livedo reticularis identifies a group of migraineurs at high risk for stroke.