Cerebellar Ataxia, Hemiplegic Migraine, and Related Phenotypes Due to a CACNA1A Missense Mutation: 12-Year Follow-up of a Large Portuguese Family

Objective To document and discuss the broad phenotypic variability in a Portuguese family with cerebellar ataxia, hemiplegic migraine, and related syndromes caused by missense mutation c.1748 (p.R583Q) in the CACNA1A gene.

Design Observational 12-year follow-up study.

Setting Community and hospital care.

Patients Sixteen patients in a 4-generation family were identified in 1998 in a population-based survey. The follow-up revealed 28 patients (25 of whom were observed) and 32 unaffected relatives with an a priori risk of 50%.

Results Four major phenotypes (migraine with multiple auras, transient focal neurological deficits without headache, coma triggered by minor head trauma, and slowly progressive cerebellar ataxia) were present in various combinations. The initial manifestation was ataxia in 16 patients and a transient episode in 12 patients. Eighteen patients did not have migraine, and 11 showed only ataxia. The c.1748 (p.R583Q) mutation in CACNA1A was confirmed in all 23 of the patients who were tested but was not found in any of the 27 adult relatives. The CACNA1A CAG repeat expansion was excluded.

Conclusions A unique missense mutation in the CACNA1A gene, which exhibits a very high penetrance and expressivity, may present a phenotypic spectrum that is broader than current descriptions. Single-gene disorders can behave as complex traits, which reinforces the importance of genetic modifiers in the tightly regulated function of P/Q-type calcium channels. The clinical spectrum of missense mutation CACNA1A -related disorders is much broader than strictly familial hemiplegic migraine.

Familial hemiplegic migraine (FHM) is an autosomal dominant type of migraine with aura (MA). Diagnostic criteria require the presence of reversible motor deficits that are associated with at least 1 other transient neurological symptom and identical episodes in at least 1 first- or second-degree relative.^1-4 The phenotype of FHM is remarkably variable,^2,3 including other paroxysmal episodes (coma,^5-7 seizures,^8-10 basilar aura,^2,11 elicited repetitive transient blindness,¹² episodic ataxia,¹³ and cerebrospinal fluid abnormalities¹⁴), permanent neurological signs (ataxia,^3,5,15,16 cognitive dysfunction,^2,17 and deafness¹⁸), and, infrequently, delayed fatal cerebral edema.⁷ Cerebellar ataxia is the most frequent of these phenotypes (up to 20% of families²) and usually begins after paroxysmal episodes^5,18,19; however, ataxia may be an isolated finding.^5,19

The first FHM gene was identified as CACNA1A (GenBank NM_001127221.1) on chromosome 19p13, which codes for the α_1A subunit of the voltage-dependent calcium channel Ca_V2.1.^13,20 Dozens of genetic mutations in CACNA1A have been described and associated with a wide range of phenotypes, including both pure¹³ and FHM-complex^5,13,21 syndromes. Half of all families with FHM (90% of those with permanent cerebellar ataxia) have a missense mutation in CACNA1A.^13,22

Furthermore, CACNA1A mutations have also been associated with allelic nonmigrainous diseases (episodic ataxia type 2, progressive cerebellar ataxia, spinocerebellar ataxia type 6, and epilepsy).⁴

We examined a family (Figure 1) with multiple cases of MA, transient focal neurological deficits (TFND) without headache, coma triggered by minor head trauma, and progressive cerebellar ataxia caused by mutation c.1748 (p.R583Q) in CACNA1A.²³ The 12-year follow-up period allowed us to refine clinical information, identify newly affected individuals, and observe rare or unknown features.

In 1998, a family doctor referred a woman and 3 children with progressive cerebellar ataxia and paroxysmal symptoms during our systematic population-based survey of hereditary ataxias and spastic paraplegias.²⁴ All of the available relatives in 4 successive generations were examined in their homes in a Portuguese fishing town. The present study was approved by the Ethics Committee of the Hospital de Santo António. After written consent, DNA was obtained from the peripheral blood of affected and healthy family members and was stored at the authorized biobank at Centro de Genética Preditiva e Preventiva, Instituto de Biologia Molecular e Celular. Genetic linkage to CACNA1A region markers was achieved, and the mutation was detected after single-strand conformation polymorphism and sequencing (Figure 2). The CACNA1A repeat expansion was detected by fluorescently labeled polymerase chain reaction and capillary electrophoresis.²³ We have been monitoring known affected patients at our clinic and observing all of the suspected and newly affected relatives. The long-term follow-up allowed us to witness the emergence of new cases, observe the temporal profile of paroxysms, and monitor the progression of ataxia and variability between generations. In 2011 and early 2012, we promoted a large clinical and laboratorial operation and updated pedigree information (Figure 1). We performed new examinations, and samples (blood or saliva) were obtained from patients and asymptomatic adults with an a priori risk of 50%, including migrants who had not been previously studied. Asymptomatic minors did not undergo genetic testing. We considered unobserved individuals as probably affected if suggestively described by at least 2 first-degree relatives.

We examined 25 patients (24 are still alive). In addition to these patients, the pedigree (Figure 1) includes 2 patients who died and 1 unreachable migrant (III:11). We also examined 34 relatives with an a priori risk of 50%. Disease was confirmed in all individuals suggestively described by first-degree relatives. We found typical symptoms in 4 subjects who were initially described as asymptomatic.

The mutation c.1748 (p.R583Q) in CACNA1A was confirmed in all 23 of the tested patients. The mutation was not found in any of the 27 tested relatives with an a priori risk of 50%: 23 were asymptomatic adults and 4 had migraine with typical aura (III:22, III:25, IV:19, and IV:23). The CACNA1A CAG repeat expansion was excluded in all of the family branches.

The Table lists the patients' sex, age at the last examination or information, and age at onset. Ten patients had MA, 11 had TFND without headache, 8 had coma triggered by minor head trauma, and 27 had progressive cerebellar ataxia (several patients had a combination of these symptoms). In addition, 1 woman (III:31) had migraine without aura. The age at the last observation (P = .25) and the age at onset (P > .99) were not significantly different between the male and female patients (Mann-Whitney test). Two individuals (III:24 and III:27) exhibited hearing loss due to homozygosity for the 35delG mutation in GJB2, which is the most common form of autosomal nonsyndromic deafness in Europe.²⁵ Detailed member-by-member characterization is provided in eTable 1, 2, and 3 .

The number of episodes, duration and type of aura, age at onset, peak periods, accompanying symptoms, precipitating factors, and other features are described in eTable 1 for each of the 10 patients with MA. There was no record of auras with hyperacute beginnings (<5 minutes). Eight patients had compound auras including motor deficit. One patient (III:27) showed isolated scintillating scotoma, and 1 patient (IV:1) had solely sensitive aura. The aura could be side locking or side shifting, even during an episode. In addition, dysphasia could appear with symptoms in any side of the body.

The age at onset and the clinical presentation, including the number of episodes, peak periods, duration and type of symptoms, accompanying symptoms, and precipitating factors, are described in eTable 2 for each of the 11 patients who had TFND without headache. Six of the patients reported 10 or more lifetime episodes. One patient (II:14) experienced a peak during her 50s and continues to have episodes at age 69 years. In one patient (IV:15), episodes recurred at age 22 years after an 11-year remission. Some of the patients displayed isolated TFND, whereas other patients had variable combinations of deficits. No patients presented with isolated visual symptoms. The majority identified recreational activities with physical stress or minor head trauma as a triggering factor.

The age at onset and the clinical presentation, including the number of episodes, peak periods, duration, and clinical features, are described in eTable 3 for each of the 8 patients with coma triggered by head trauma. All of the episodes were precipitated by minor head trauma without visible head damage and involved a negative or inconclusive investigation. Five patients reported a single episode. Two early-onset patients (IV:9 and IV:10) described several episodes of coma. Childhood play and sports were the most frequent causes of head trauma. Some episodes were accompanied by hemiplegia (II:11, III:37, and IV:10).

Twenty-seven patients had permanent, slowly progressive cerebellar ataxia. The patients' ages at onset are presented in the Table and Figure 3. Except for one 20-year-old woman (IV:16), all of the examined patients had ataxic gait with variable severity; however, all of the patients were able to maintain some autonomy: 1 patient (II:11) had progression over 33 years using 1 crutch, 2 patients (I:2 and II:14) used occasional support against the wall while walking, 14 patients had staggering gait and difficulty on half turns, and the remaining patients were unable to perform tandem walking. Limb ataxia was present in 13 patients, and 15 patients had dysarthria. In addition, 3 patients presented with horizontal gaze nystagmus. Cerebellar vermis atrophy was found in 6 of 7 patients who underwent brain magnetic resonance imaging. Interestingly, one 39-year-old woman (III:14) with ataxia since age 14 years had normal findings on magnetic resonance imaging.

The 3 unobserved patients were described as having isolated ataxia. In the 25 patients who were examined, the 4 main syndromes were expressed in various combinations to produce 9 different phenotypes (Figure 1 and Table). Nine patients had isolated syndromes (8 had ataxia and 1 had TFND). Four patients had the 4 main syndromes. Among the 16 patients with 2 or more syndromes, the first manifestations were variable (5 ataxia, 5 TFND, 3 MA, 2 coma, and 1 TFND/coma). Ataxia and episodic symptoms began in the same year in 3 patients. One patient (II:9) had ataxia since age 40 years, and she had a transitory episode of aphasia and right hemiplegia that lasted 8 hours at age 74 years; vascular investigation and cerebrospinal fluid analysis findings were normal.

All of the patients in the third generation of branch A had isolated ataxia (Figure 1 and Table). Some of their children had early ataxia accompanied by numerous paroxysmal episodes. Interestingly, the mean (SD) age at onset decreased from 33.5 (7.2) years to 9.3 (5.3) years between 2 generations. In branch B (mean [SD] age at onset, 19.5 [10.9] years), MA was never observed. Branch C (mean [SD] age at onset, 16.0 [8.7] years) had patients with 3 different syndrome combinations. In branch D (mean [SD] age at onset, 13.6 [5.9] years), there were different combinations in each generation, although none had isolated ataxia.

We describe a family segregating an autosomal dominant variable phenotype due to mutation c.1748 (p.R583Q) in CACNA1A whom we followed up for a 12-year period. The mutation was confirmed in all 23 patients tested and was not found in any of the 27 tested relatives with an a priori risk of 50%; thus, the mutation exhibited a very high penetrance and expressivity. The CACNA1A CAG repeat expansion was excluded.

This is most likely the largest family with FHM type 1–related disorder ever reported and the family with the widest phenotypic variation. Interestingly, no patients experienced some of the previously reported FHM-related syndromes such as basilar migraine,^2,11 cognitive impairment,² or epilepsy.^9,10

Screening for FHM based on relatives' testimonials was specific but not sensitive. The diagnosis of FHM type 1 based on the first-degree relatives' information seems reliable because all of the patients who were described as symptomatic were confirmed to have the disorder; however, exclusion of the diagnosis on the same grounds cannot be considered because some subjects described as asymptomatic were found to be affected after our examination. Ataxia was the most frequent phenotype and was almost universal in adults. Chronic progressive cerebellar ataxia appeared after paroxysmal episodes (9 patients), in advance of these (5 patients), or as an isolated feature (11 patients); this confirms and adds robustness to descriptions of the variable natural history of ataxia in FHM type 1–related disorders.^5,18,19,23 The majority of our patients, however, did not have nystagmus, which disagrees with previous reports. One patient (III:14) with a long history of ataxia had normal findings on brain magnetic resonance imaging, which suggested that atrophic cerebellar degeneration is not a necessary pathological expression of channelopathies.

The number and temporal profile of paroxysmal transient episodes were extremely variable. Syndromic combinations and age at onset were also rather variable. Paroxysmal episodes were more frequent in the second or third decades and tended to attenuate with age, whereas ataxia worsened with age. Only 10 of the 25 observed patients had MA, and only 8 of these 10 patients had hemiplegic aura. This differs from most reported series of CACNA1A -related disorders. Although aura varied in type, location, body side, and topographic congruence (eg, aphasia with right or left hemiparesis), the duration was identical to MA but shorter than the duration usually described in FHM.² Headache was migrainous in character but mostly bilateral. Cumulative 10-year prevalence of migraine without aura in this family was lower than in the general population.²⁶ This may suggest that migraine without aura and calcium channelopathies are unrelated, which has previously been proposed.²⁷

Six of the 11 patients with TFND without headache also had MA, with similar age at onset for both syndromes. Although some features of TFND are similar to aura, identified differences were lack of scintillating scotoma or isolated visual aura, longer duration, and more frequent precipitating factors. Some patients with TFND had early remission, whereas 1 had an apparent remission and another had the first episode at age 74 years. These findings may suggest that there is a higher triggering threshold for TFND, which may require multiple and stronger endogenous and environmental stimuli. Episodic coma triggered by minor head trauma occurred in patients with a younger age at onset and was often a single event. The predominance of males could be explained by a greater exposure to coma triggers during childhood. The frequent association of coma with TFND in the same episode and shared precipitating factors suggest a common pathogenesis.

The migrainous nature of sensorineural deafness, which was found in 3 of our patients and previously described¹⁸ in FHM, was excluded by genetic analysis. We recommend judicious evaluation of atypical symptoms before assuming that they are part of the channelopathic spectrum.

All of the patients of the third generation in branch A had isolated cerebellar ataxia. Disease onset was earlier in the fourth generation, and the phenotype was more severe and complex. The tendency for anticipation of age at onset seems real; however, it is generally agreed that true anticipation is mimicked by random variation in severity and ascertainment biases. The absence of paroxysmal symptoms in the third generation can be attributed to a recall bias, which has been documented for sporadic hemiplegic migraine.²⁸ A lifelong history of isolated cerebellar ataxia does not preclude the later onset of focal episodes (eg, patient II:9).

It is widely accepted that mutations of CACNA1A cause a variety of manifestations depending on differing underlying mutations. In the present family, 1 missense mutation manifested as a syndromic continuum with an even broader and more heterogeneous spectrum than previously reported. Functional studies will help to explain how the variability can be attributed to a single mutation.

Electrophysiological studies have shown that the R583Q mutation shifts the voltage dependence of activation to more hyperpolarized potentials and slows channel recovery from inactivation.²⁹ These alterations in channel properties result in increased P/Q currents (even at weak depolarizations) that have been suggested to be the cause for the initiation of cortical spreading depression as a result of minor head trauma. In addition, increased calcium influx probably results in neuronal dysfunction and Purkinje cell death and thus correlates with the ataxic features found in patients with the R583Q mutation. Moreover, the clinical variability found in the present patients with the R583Q CACNA1A mutation is a good example of how single-gene disorders can behave as a complex trait and reinforces the importance of genetic modifiers in the tightly regulated function of P/Q-type calcium channels.

The clinical spectrum of CACNA1A -related disorders is much broader than strictly FHM. A family history of hemiplegic aura, coma triggered by minor head trauma, and progressive cerebellar ataxia should be accepted as diagnostic evidence for FHM. In contrast, migraine with typical aura does not predict a diagnosis of channelopathy, even in individuals from a susceptible family. Perhaps this statement could be considered in the future review process of the International Classification of Headache Disorders.¹

Correspondence: José Barros, MD, Serviço de Neurologia, Hospital de Santo António, Centro Hospitalar do Porto, Largo Professor Abel Salazar, 4099-001 Porto, Portugal (josebarros.neuro@hgsa.min-saude.pt).

Accepted for Publication: April 5, 2012.

Published Online: November 12, 2012. doi:10.1001/jamaneurol.2013.591

Author Contributions:Study concept and design: Barros and Coutinho. Acquisition of data: Barros, Damásio, Tuna, Alves, and Coutinho. Analysis and interpretation of data: Barros, Damásio, Tuna, Silveira, Pereira-Monteiro, Sequeiros, Alonso, Sousa, and Coutinho. Drafting of the manuscript: Barros, Damásio, Sequeiros, Alonso, Sousa, and Coutinho. Critical revision of the manuscript for important intellectual content: Barros, Tuna, Alves, Silveira, Pereira-Monteiro, Sequeiros, Alonso, Sousa, and Coutinho. Statistical analysis: Sousa. Obtained funding: Barros, Silveira, Pereira-Monteiro, Sequeiros, Alonso, and Coutinho. Administrative, technical, and material support: Barros, Damásio, Tuna, and Alves. Study supervision: Barros and Coutinho.

Funding/Support: This study was supported by grants PRAXIS /P/SAU/13226/1998, POCTI/32643/ESP/2000, and PIC/IC/83232/2007 from the Fundação para a Ciência e Tecnologia, the Tecnifar award and grant from the Sociedade Portuguesa de Cefaleias, and the SPN award and grant from the Sociedade Portuguesa de Neurologia. Dr Alonso is funded by the Programa Ciência, POPH-QREN–Tipologia 4.2–Promoção do Emprego Científico, is cofunded by the European Social Fund, and has received national funds from the Ministério da Ciência e Ensino Superior.

Additional Contributions: Graça Fialho, PhD, and Helena Caria, PhD, Faculdade de Ciências, Universidade de Lisboa, performed the molecular study of congenital deafness, Teresa Gomes performed blood sampling, and Centro Hospitalar Póvoa de Varzim/Vila do Conde allowed the use of their facilities.