The Endless Expansion of the Phenotypic Spectrum of ATP1A3 Mutations: A True Diagnostic Challenge

With a relentless appearance in the medical literature of reports describing novel clinical presentations, the spectrum of phenotypes associated with mutations in ATP1A3, encoding the sodium/potassium‐transporting ATPase subunit α3, continues to expand. Although this represents a formidable example of the diagnostic revolution deriving from the advent of modern sequencing technologies, the drawback is that further complexity is constantly added to an already convoluted genotype‐phenotype correlation. The identification of ATP1A3‐related conditions may represent a true diagnostic challenge even for experienced clinicians.

Initially reported as the cause of rapid‐onset dystonia‐parkinsonism (RDP), a very rare autosomal dominant condition characterized by abrupt onset in adolescents or young adults of fixed dystonic and parkinsonian symptoms,1, 2 it was subsequently recognized that de novo ATP1A3 mutations are also responsible for alternating hemiplegia of childhood (AHC).3, 4 AHC is an infantile disorder characterized by a fluctuating course and by recurrent and alternating, hemiplegic or hemidystonic episodes, followed in most cases by the development of permanent neurological symptoms (i.e., epilepsy, neurodevelopmental delay with intellectual disability, chorea, dystonia, parkinsonism, bulbar symptoms, and cerebellar ataxia).5 More recently, it was demonstrated that a single missense mutation causes another distinct autosomal dominant condition, featuring progressive cerebellar atrophy with relapsing episodes of cerebellar ataxia, areflexia, pes cavus, optic atrophy, and sensorineural deafness (CAPOS).6 In their most classic descriptions, the three phenotypes—RDP, AHC, and CAPOS—differ in many aspects, including main neurological features, range of age at onset, and disease course. Furthermore, initial observations seemed to indicate there was no or little mutational overlap, suggesting that AHC, RDP, and CAPOS were likely to be individual allelic entities.4

In this issue of Movement Disorders Clinical Practice, two articles by Pereira and colleagues7 and de Gusmao and colleagues8 remind us that, unfortunately, things are not this simple.

Pereira and colleagues describe a child who presented at age 3 years with acute dystonia affecting the left upper limb, which was followed in the next few days by severe, fluctuating (aggravated by exercise and improved by sleep) bulbar symptoms and action tremor on the contralateral side. Interestingly, approximately 3 months after the initial presentation, he also developed paroxysmal episodes of alternating hemiparesis, hypotonia, and bulbar symptoms, fully consistent with the episodes usually observed in AHC.

It is important to note that the clinical features of this case, which largely overlapped between AHC and RDP, did not meet the current diagnostic criteria for the diagnosis of either condition. Although the initial presentation would possibly fit the description of RDP,2 the age at onset was unusually early. Furthermore, the overtly fluctuating course and the presence of an action tremor are atypical for classic RDP (the absence of tremor at onset is one of the suggestive diagnostic features). Again, the age at onset of the AHC symptoms was atypical, as the diagnostic criteria for AHC indicate that the symptoms should begin before the age of 18 months.9

Consistent with these observations, a growing number of cases (some of them summarized by Pereira et al.) with overlapping RDP/AHC10, 11, 12, 13, 14 and CAPOS/AHC15 have been reported in the literature, suggesting that intermediate phenotypes may actually be far more common than initially thought.

The work of de Gusmao and colleagues raises a second important aspect. The clinical presentation of ATP1A3 mutation carriers may not encompass the full spectrum of symptoms usually observed in AHC, RDP, or CAPOS, but only a subset of features.

Their Case 3 is particularly instructive in this regard. A 28‐year‐old man developed progressive ataxic syndrome in his 20s accompanied by myoclonus and action tremor. His brain magnetic resonance image (MRI) showed evidence of vermian cerebellar atrophy. Importantly, there was no history of a clear trigger preceding the onset, and other features observed in patients with CAPOS were absent. This case indicates that ATP1A3 mutations should be considered when assessing patients who have genetically unsolved cerebellar ataxia; progressive myoclonus‐ataxia syndrome; and, given the paroxysmal nature of the ATP1A3 mutation‐related disorders, episodic ataxia. It is interesting that the main neurological feature of case 1, whose symptoms fell within the RDP spectrum (bulbar symptoms started after a febrile illness), was generalized chorea mixed with dystonic elements. Hence, it certainly will be important to suspect ATP1A3 mutations in the differential of genetically undetermined choreic syndromes.

A range of other atypical ATP1A3‐related presentations have been described, ranging from milder presentations (e.g., levodopa‐responsive paroxysmal oculogyria and mild neck posturing)12 to more aggressive ones (e.g., catastrophic epileptic encephalopathy and microcephaly or relapsing encephalopathy associated with progressive cerebellar ataxia).16, 17 Furthermore, patients with ATP1A3 mutations also can display significant neuropsychiatric features, including psychotic episodes,18 attention deficit and hyperactivity disorder (as in one of the patients described by de Gusmao et al.), and cognitive impairment.19

Overall, the phenotypic spectrum secondary to ATP1A3 mutations is much broader than initially observed, and it is highly likely that the spectrum of neurological, and possibly psychiatric, presentations will continue to grow as more patients with genetically unsolved disorders will undergo next‐generation sequencing analysis.

In this scenario, from a clinical point of view, one of the outstanding questions is how clinicians should suspect ATP1A3 mutations when evaluating patients who have such unusual presentations. As Pereira and colleagues point out the strict use of the classical diagnostic criteria for AHC, RDP, or CAPOS may delay diagnosis in a significant proportion of patients, suggesting that the threshold for suspecting ATP1A3 should be lowered.

Although (1) a paroxysmal nature of symptoms, (2) an abrupt presentation and/or a rapid deterioration, and (3) a prominent bulbar involvement are very commonly observed, clinicians should be aware that these suggestive features might not always be present (as eloquently demonstrated by case 3 in the report by de Gusmao et al.). Moreover, the disease course and the presence of triggering factors seem to be highly variable as well, an element that makes it even more difficult to suspect ATP1A3 mutations on clinical grounds alone.

With more and more atypical phenotypes reported in the literature, it will be essential to carefully recognize and define the key red flags for the clinical suspicion of ATP1A3 mutations that should prompt genetic testing.

It is important to note that the identification of ATP1A3 mutations bears crucial therapeutic and prognostic implications. First, the administration of flunarizine, a channel blocker well known to reduce the frequency and intensity of plegic attack, may also be effective in preventing motor deterioration in patients with AHC.20 Although this observation awaits confirmation in randomized trials, it is possible that the use of flunarizine also may protect patients with other ATP1A3‐related presentations against motor deterioration.

Furthermore, faulty cardiac repolarization has recently been demonstrated in ATP1A3 mutation carriers, and this could account for the reported increased rate of premature mortality due to cardiorespiratory arrest.21 Therefore, a prompt diagnosis is important in all patients with ATP1A3‐related conditions, because a systematic cardiac investigation may significantly reduce the risk of this catastrophic, but preventable, outcome.

Disclosures

Funding Sources and Conflicts of Interest: Dr. Mencacci received support from the Department of Health's National Institute for Health Research (NIHR) Biomedical Research Centers funding streams. He reports no conflicts of interest.

Financial Disclosures for the previous 12 months: The author reports no sources of funding and no conflicts of interest.