Introduction

Cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL) is an adult-onset inherited disease due to mutations in the Notch3 gene on chromosome 19p13 [1–3]. These mutations cause an abnormal accumulation of Notch3 in the cytoplasmic membrane of vascular smooth muscle cells both in cerebral and extracerebral vessels, which appear as granular osmiophilic deposits on electron microscopy [4]. The neuroradiological hallmark of the disease is represented by leukoencephalopathy with lacunaes in the basal ganglia, which can become evident at any stage of the disease [4]. The clinical spectrum of CADASIL includes migraine with or without aura, mood disturbances, transient ischemic attacks or strokes (usually lacunar infarcts) and progressive cognitive decline [1, 2, 4, 5]. Other disturbances are also reported, including epilepsy, acute reversible encephalopathy and myopathy [6, 7].

Headache in patients with CADASIL was first coded, according to the Classification and Diagnostic Criteria for Headache Disorders, Cranial Neuralgias and Facial Pain (ICHD), among headaches associated with other vascular disorders (code 6.9) [8]. Following the ICHD, second edition (ICHD-II), headache in patients with CADASIL is coded among the headaches attributed to cranial or cervical vascular disorder (code 6.7.1) on the implicit assumption that a vascular disorder is the cause of the headache [9], an interpretation that, in our opinion, should be reconsidered [5, 10].

To verify the appropriateness of the ICHD-II criteria, we deemed it to be of interest to review the studies reporting any headache in patients with CADASIL.

Methods

Available data for this review were found through Medline/PubMed using the keyword “cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL)”. For our study, we reviewed all studies published in English between 1993 and 2008. All the retrieved studies were examined by the two of us (SS and DD). We excluded studies that did not report original data on CADASIL such as reviews, meta-analyses, comments and editorials. Among studies describing clinical details of patients with CADASIL, we excluded those in which there was no mention of headache or migraine. To exclude overlapping of CADASIL cases by the same research group, we considered what authors wrote in their studies and, where necessary, contacted the corresponding author.

Two of us (SS and DD) evaluated and classified all the reported cases separately. Any possible disagreement was resolved by a third opinion (AC). We maintained the terminology used by the authors in their studies and, consequently, where authors classified the headache according to the ICHD-II we kept their classification and where they used terms not contemplated in the ICHD or ICHD-II we kept their terminology. We also considered whether the headache was the first symptom or whether it was preceded by an ischemic stroke, TIA or other symptoms such as cognitive and psychiatric disturbances.

Mean age ± standard deviation (SD) was calculated with the method of weighted mean.

Results

The Medline/PubMed search identified 590 studies; 508 studies were excluded because they did not report clinical data or because there was no mention of headache, and 48 because their data had been already included in other studies by the same authors. We finally reviewed 34 (5.8%) studies including 16 case reports or letters (Table 1) [1, 5–7, 11–40].

Table 1 Studies included in the review
Full size table

The reviewed studies provided data on 749 patients including 387 (51.7%) who had headache. According to the authors’ definition, 356 (92%) patients were reported as having migraine and 31 (8%) as having headache (Table 2).

Table 2 Overall characteristics of patients with CADASIL
Full size table

Out of the 356 patients described as migraineurs, 125 (35.1%) had migraine with aura, 7 (2%) migraine without aura, 156 (43.8%) unspecified migraine, and 68 (19.1%) had more than one type of migraine (Table 3). Among the 125 patients who had migraine with aura, 24 (19.2%) had migraine with typical aura, 7 (5.6%) migraine with prolonged aura, 5 (4.0%) familial hemiplegic migraine, 2 (1.6%) basilar migraine, while in 87 (69.6%) patients the migraine type was not reported. The aura was visual in 10 (8%) patients, sensory in 7 (5.6%), motor in 5 (4%) and included multiple symptoms in 25 (20%); the aura type was not specified in 78 (62.4%) patients. Among the 68 patients with more than one type of migraine, 50 (73.5%) had migraine with aura, 38 (55.9%) migraine without aura, 18 (26.5%) migraine with aura and without headache, 13 (19.1%) basilar migraine, 11 (16.2%) familial hemiplegic migraine, 11 (16.2%) migraine with unspecified atypical aura, 4 (5.9%) acute onset aura without headache, 2 (2.9%) unspecified migraine, 1 (1.5%) migraine with prolonged aura, 1 (1.5%) migraine with acute onset aura, and 1 (1.5%) had status migrainosus (Table 3). At the onset of migraine, the mean age ± SD was 32.2 ± 13.1 years.

Table 3 Distribution of migraine subtypes in patients with CADASIL
Full size table

In the 31 patients reported as having headache, the headache was not further specified in 18 (58.1%) patients, was defined as chronic in 6 (19.3%), as resembling migraine with aura in 4 (12.9%), as resembling migraine without aura in 2 (6.5%), and as tension type headache in 1 (3.2%) patient. At the onset of headache, mean age ± SD was 38.9 ± 16.8 years.

We also analyzed whether the headache was the first symptom or whether it was preceded by an ischemic stroke, TIA or by other symptoms. After excluding patients for whom data were not reported, 81 patients were described as suffering from migraine and 19 from headache. While 9 of the 81 patients (11.1%) had a previous stroke or TIA, migraine was the first symptom in 66 patients (81.5%) and in 6 (7.4%) it was preceded by other symptoms. While 3 of the 19 patients (15.8%) had a previous stroke or TIA, headache was the first symptom in 14 (73.7%) and in 2 (10.5%) it was preceded by other symptoms.

Discussion

From the 34 reviewed studies, we collected data on 387 patients [1, 5–40]. In none of the cases, the headache was coded according to the ICHD [8] or ICHD-II [9] criteria. The ICHD-II diagnostic criteria for Headache attributed to CADASIL required the following: (A) attacks of migraine with aura, with or without other neurological signs; (B) typical white matter changes on magnetic resonance imaging T2-weighted sequences; (C) diagnostic confirmation from skin biopsy evidence or genetic testing (Notch3 mutations) [9]. The coding of headache according to the ICHD or ICHD-II criteria might not have been done due to the paucity of details reported in the previous and present classification. Had authors correctly applied the ICHD-II criteria, patients with CADASIL would have been diagnosed as suffering from headache attributed to CADASIL without any further detail. In the reviewed studies, we found that 51.7% of patients with CADASIL were reported to have suffered from headache. Anyhow, despite that headache represents one of the major clinical features of CADASIL, we cannot exclude that some patients suffered from a primary headache. Besides, based on age of onset and clinical characteristics of the headache in CADASIL resembling those of some primary headaches, it was not possible to establish the proportion of primary headaches in the same patients. The headache in patients with CADASIL was more frequently and clearly defined as migraine and the majority of the migrainous patients were reported as suffering from migraine with aura (35.1%). Formally, this codification was inappropriate since diagnostic criteria for any type of migraine in the ICHD-II require that the disturbance is not attributed to another disorder. If the classification of CADASIL in the ICHD-II was not revised, clinicians will be induced to commit a formal error defining the headache as migraine or omit useful clinical details attributing the headache to CADASIL without giving any further characteristic. As much as 31 cases were diagnosed with disturbance other than migraine; in those cases, the authors used a generic terminology and possibly included patients suffering from primary headaches such as tension-type headache.

The inclusion of headache in CADASIL, in chapter 6 of ICHD-II, among Headache attributed to cranial or cervical vascular disorder, implies that the headache is caused by a vascular disease. This implication may rely on two possibilities: the former is that the headache in CADASIL might be considered secondary to the presence of an organic vascular lesion and the latter is that a disorder of the vascular system might be considered as the common underlying pathogenic mechanism that causes both headache and stroke. Referring to the former possibility, in CADASIL, the headache is usually the first symptom of the disease (73.7–81.5% according to reviewed data) preceding the onset of stroke or TIA. Moreover, a diagnosis of secondary headache is usually evident only when the headache resolves or greatly improves within a specified time interval after its onset or after the acute phase of the vascular disorder; however, this evolution is not reported in patients with CADASIL since the headache persists across the years. Consequently, for all the above reported reasons, this possibility is unlikely. Referring to the latter possibility, we have to consider that the underlying pathology in CADASIL is represented by an angiopathy with a unique type of ultrastructural basal lamina deposits and by degeneration of vascular smooth muscle cells, which are the major source of the Notch3 expression. The evidence for a functional impairment of vascular smooth cells is in line with this latter hypothesis [41] and consequently the pathogenic assumption reported in the ICHD-II is correct. However, we would underscore that the primary mechanism is not represented by the vascular damage, but by the genetic alteration in the Notch3 expression. For all these reasons, we suggest considering the possibility of revising the ICHD-II when referring to CADASIL. Specifically, we suggest adding a new category that could be named Headache attributed to genetic disorder including Headache attributed to CADASIL (Table 4). The new category might include also other genetic diseases in which headache represents a major clinical feature. Moreover, the new category should also report subtypes according to specific clinical details to allow the precise characterization of headache in the single patient (Table 4). The proposed diagnostic criteria for headache attributed to CADASIL are reported in Table 5. If our proposal is shared by the experts in the field, we think that it shall be easier from now on to characterize patients with CADASIL uniformly and that a good step forward will be realized.

Table 4 Proposal for a new classification of headache in patients with CADASIL based on available data
Full size table
Table 5 Proposed diagnostic criteria for the new category of headache attributed to genetic disorder
Full size table