Use of next-generation sequencing, including whole-exome sequencing (WES) has not only allowed diagnosis to be reached in patients with atypical phenotypes, but also led to detection of new pathogenic variants, as well as to linking of specific clinical manifestations to known diseases. Nevertheless, controversy persists regarding routine implementation of WES in clinical practice [1, 2]. Interpretation of results and understanding of the clinical relevance can be problematic, particularly in cases of variants of unknown significance (VUS) [3]. Application of WES can however be cost-effective when based on appropriate clinical criteria, and may even reduce costs by limiting unnecessary complementary studies [4]. Although slowly becoming more financially accessible, WES remains an expensive study. Inappropriate or indiscriminate use may increase overall healthcare costs, without providing significant benefit. In addition, WES does not detect large gene deletions or duplications, or expansion disorders such as triplet repeat expansions, or genes located on non-coding segments of the genome (introns), and therefore is not useful to diagnose diseases caused by these particular mutations [1, 2]. The diagnostic yield of WES in different case series of adult patients with neurological diseases has consistently been around 30% [5, 6]. In children, similar [7, 8, 9], or somewhat higher values have been reported [10]. More specifically in movement disorders, a study including 378 patients with atypical or combined phenotypes, found the diagnostic yield was 22% [11].

In this context, we set out to analyze the indications for WES, through the analysis of results obtained in a cohort of 2948 patients consulting the Movement Disorders clinic at our institution in Buenos Aires, Argentina between July 2015 and July 2019. Prior to indicating the test, all patients had undergone thorough clinical evaluation by a movement disorder specialist, as well as routine diagnostic workup to exclude frequent disorders treated at our tertiary clinic. WES was ultimately indicated based on patient phenotype in 54/2948 (1.8%) patients, lacking definitive diagnosis after exhaustive clinical examination as well as standard blood, urine, and imaging studies, including selective genetic testing for Huntington’s disease, Friedreich’s ataxia, spinocerebellar ataxias due to repeat expansions and DYT1 gene mutations. WES was also requested in cases of combined or atypical phenotypes presenting ataxia, parkinsonism, chorea, tremor, dystonia and spastic paraplegia, and/or patients with positive family history, or early onset of disease.

Of the 54 patients in which WES was obtained,14 (25.9%) presented pathogenic variants and 18 (33.3%) VUS. In 27 (50%), genetic tests for common repeat expansions inducing movement disorders were negative. Main clinical and genetic characteristics are summarized in Table 1. Similar to other results reported in the literature on the diagnostic yield of WES [5, 6], in this series of 54 patients, 17 variants were considered clinically pathogenic (35.2%). TRIO analysis was conducted in 7 of 32 patients with genetic findings (21.9%); and proved useful to confirm pathogenicity, in cases where compound heterozygous mutations were present (cases 22, 28, 29). In case 31, in which no family history of autosomal dominant dystonia was detected, TRIO analysis led to a diagnosis in the mother, rather than confirmation of pathogenicity of the mutation. The number of variants interpreted as pathogenic according to phenotype were: 7/19 (36.8%) in ataxia; 4/18 (22.2%) in parkinsonism; 3/3 (100%) in tremor, 1/7 (14.3%) in spastic paraplegia and 6/21 (28.6%) in dystonia. In two patients presenting chorea, no pathogenic variants were identified.

Establishing etiology enabled starting or modifying evidence-based treatment in 9 patients, in whom the variant identified was either considered pathogenic, or clinically interpreted as disease-causing. Treatments included conservative measures such as avoiding triggers in DYT12 cases [12]; prescribing supplements such as chenodeoxycholic acid in cerebrotendinous xanthomatosis [13], docosahexaenoic acid in SCA38 [14], acetazolamide in the congenital disorder of glycosylation Ia (CDG-Ia) due to mutations in the PMM2 gene [15]; as well as use of trihexyphenidyl for tremor treatment in juvenile Parkinson’s disease due to PRKN mutation [16]. In both patients presenting pathogenic variants in the PMM2 gene, and in one patient with POLG mutation, WES results allowed expansion of the phenotypic spectrum of the disorders, since dystonia had not been previously reported in the literature for these particular conditions [17, 18]. For more information detailing treatments and outcomes, please see Supplementary Table 1.

In one patient who was adopted, family history was not available. Of the remaining 53, 12 had positive family history (22.6%). Of these, 9/32 (28.1%) presented positive test results combined with a positive family history, and 3/22 (13.6%) had a negative test result.We found no significant correlation between positive test result and age at disease onset or positive family history, probably due to the limited total number of patients. However, correlation was observed between positive test results and tremor phenotypes.

Major limitations to this analysis include the retrospective nature of the data collection, small sample size, lack of standardized diagnostic algorithm prior to WES, and heterogeneous criteria applied for indicating WES beyond treating physician preference, potentially generating significant selection bias as to which patients were tested and which were not. Furthermore, few patients were able to afford the study, therefore not all WES studies prescribed were performed. Because this was a consecutive cohort of patients studied with WES, and not a consecutive cohort of patients with atypical or combined phenotypes, diagnostic yield of WES could not be reliably assessed.

The ideal scenario in which to indicate WES is still hard to define. As a diagnostic tool it is still in early stages of development, with evolving discoveries, and only recently becoming more readily accessible. Further cohort studies may help determine which specific clinical characteristics will yield better test results. Currently, we believe WES should be considered early on, in patients with movement disorders presenting atypical or combined phenotypes, or early age at onset and positive family history, or in those lacking a clear diagnosis following thorough clinical evaluation by a trained specialist. The significant gap (Table 1) between symptom onset and definitive diagnosis based on WES, suggests earlier use could shorten time to start of treatment, as well as dampen patient diagnostic uncertainty.

Conversely, misinterpretations or VUS may decrease diagnostic accuracy. and lead to unnecessary testing or therapeutic trials, all of which can be considerably reduced when patients are evaluated by trained specialists.