Genetic analysis of a diagnosed ICC is highly recommended to unravel the cause of the disease (Priori & Blomstrom-Lundqvist, 2015). A high percentage of patients carrying a deleterious genetic variant remains asymptomatic, thus, early identification of a genetic alteration definitively associated with the disease may drive therapeutic pharmacological and non-pharmacological management to reduce risk of lethal arrhythmias (Ackerman et al., 2011, Campuzano, Sarquella-Brugada, Arbelo, et al., 2020, Musunuru et al., 2020). However, a main challenge for the clinical translation of genetic data is understanding the definite role of a rare variant. Misinterpretation of rare variants may lead to inaccurate genetic diagnoses and/or the adoption of unnecessary and/or inappropriate preventive and therapeutic approaches.
Currently, the classification of a rare variant is performed following the ACMG recommendations and according to data available at the moment of interpretation (Richards et al., 2015). A geneticist who misinterprets the significance of a variant can be considered legally liable only if the current scientific evidence and the application of these criteria whould have permitted a correct classification (Grassi et al., 2020). However, a misinterpretation is often due to the lack of sufficient scientific evidence. Continuous development of clinical data, bioinformatic tools and advances in genomic knowledge highlights the critical need for periodic revision (Salfati et al., 2019). In addition, publication of new data may also alter the prior classification of a rare variant, but there is, as yet, no concrete timeframe for reanalysis. Our study shows that nearly 20% of overall variants changed classification within 5 years, suggesting a lustrum should be the potential minimum time for reanalysis, at least for rare variants associated with main ICC.
Impact of reanalysis
The same number of rare variants with a deleterious/potentially deleterious role were identified in 2016 and now (30.61%), and all located in main genes associated with ICC (SCN5A, KCNQ1 and KCNH2). One variant upgraded from LP to P, accordingly to a recent manuscript by Harrison et al. that found that LP variants became P between 2016 and 2019 (Harrison & Rehm, 2019). However, the results of the Harrison et al. study were based on variant modifications identified only in ClinVar, with the authors being uncertain that all classifications were according to ACMG guidelines. The authors did state that most radical variants, i.e., those leading to premature truncation of proteins and/or frameshifts, should be considered highly damaging if not identified at high frequency in population and, therefore, should be carefully analyzed. Our results concord with this point as the majority of indels in our study remained classified as LP except one, showing high frequency in global population and classified as LB. Therefore, for each patient, missense rare variants should be comprehensively analyzed considering all available data to be properly prioritized in a personalized clinical context (Eilbeck, Quinlan, & Yandell, 2017).
In the case of VUS, a rare variant classified as ambiguous does not provide molecular confirmation of a diagnosis but cannot be discarded as indicating a low risk of malignant arrhythmias for any patient, at least until additional data clarifies its clinical role (Musunuru et al., 2020). In our study, most rare variants were classified as VUS in both periods of analysis (from 69.38% to 53.06%, currently). All VUS changed to LB due to substantial increase of MAF seen with ongoing analysis of the global population. This point notes the key role of global frequencies and its correlation with the prevalence of the ICC in the population. Thus, reducing uncertainty by excluding the potential effect of some VUS variants, at least on ICC, is probably one of the remarkable benefits of reanalysis. These results accord with those of previous studies showing VUS rarely change to P or LP variants (Lahrouchi et al., 2017). However, until the role of a rare variant can be certainly or likely assessed, no inference should be made on its possible significance and no impact in clinical management of the patient can be justified. Therefore, clinical translation of VUS should be performed with caution and families counselled regarding the current limitations of reliable clinical interpretation. In our opinion, a complete family segregation together with an accurate global frequency are the most robust tools to distinguish a potential damaging variant from other rare variants, and free access to available frequency databases is a quick and crucial step. Other data, e.g., functional studies, may corroborate the pathogenic role of a rare variant (Glazer et al., 2020), but unfortunately, a complete family segregation as well as functional analysis are not available for most rare variants currently associated with ICC. It is also important to note that a low percentage of current VUS may eventually be shown to confer a real risk of ICC (Blekhman et al., 2008), however, distinguishing pathogenic VUS from the majority of VUS is one of the main current challenges in the ICC field (Cherny, Olson, Chiodo, Balmert, & Webster, 2020, Tsai et al., 2019).
The ACMG recommends how variants should be classified, but there is no consensus for how often reclassification should occur. Therefore, reanalysis of rare variants occurs mainly due to a clinician’s request, identification of a previously classified variant in a new patient, or other new data (El Mecky et al., 2019). Smith et al. reported that 3% of ICC rare variants became reclassified after one year.(Smith et al., 2017) In another study, reclassification after 7 years (from 2011 to 2018) increased the molecular diagnosis of ICC by 2% in cases of unexpected decease (Salfati et al., 2019). In an additional cohort analyzed for genetic arrhythmia and cardiomyopathy disorders from 2006-2017, 22% of variants changed classification and approximately 10% changed in a way that altered clinical interpretation (Cherny et al., 2020). Following similar results, VanDyke et al published that 35% of the variants associated with ICC had classifications that differed from their firsts reports (all prior 2015) (VanDyke et al., 2020). Therefore, it is widely accepted that variants did not classify following ACMG recommendations should be updated immediately due to the potential clinical impact associated (Campuzano, Sarquella-Brugada, Fernandez-Falgueras, et al., 2020). In addition, it is also accepted that periodic reinterpretation of variants is necessary for clinical management of patients with ICC but remains undefined how often a reanalysis is necessary if variants are already classified following ACMG recommendations. Our study provides first evidence that a 5-year timeframe is adequate to manage the rapid obsolescence of genetic data interpretations, at least in ICC.
Brugada syndrome
In 2019, Denham et al reported that only 37% of rare variants previously considered deleterious for BrS were definitely reclassified as P or LP following ACMG guidelines.(Denham et al., 2018) Accordingly, in our study 23.52% of BrS rare variants were classified as deleterious in 2016, being currently the same percentage classified with a harmful role. All these rare deleterious variants were located in SCN5A. The number of VUS was reduced since 2016 (from 13 to 7), 6 to LB (in minor genes) and only one to LP (in SCN5A). Taking all these data into account, our study reinforce SCN5A as the major gene associated with BrS (Walsh & Wilde, 2019).
Catecholaminergic polymorphic ventricular tachycardia
In our cohort of CPVT, all identified rare variants were located in RyR2, reinforcing the role of this gene as the major player in CPVT. A recent study reclassified all RYR2 variants using ACMG recommendations and found that only 7.8% of previously disease-associated variants remained as LP/P (Olubando et al., 2020). Another study published in 2018 found that five upgraded from VUS to LP/P and six downgraded from P/LP to VUS when ACMG recommendations were applied (Roston et al., 2018). The same trend was observed in a cohort of children in which two out of five (40%) VUS in RYR2 upgraded to LP/P using ACMG recommendations (Bennett et al., 2019). This supports an assertion made in prior studies where an urgent reanalysis of rare variants associated with ICC should be done if they were not classified originally following ACMG recommendations (Campuzano, Sarquella-Brugada, Fernandez-Falgueras, et al., 2020, VanDyke et al., 2020). The number of VUS in our cohort of CPVT cases did not change from 2016 to date. In comparison to BrS and LQTS, the estimated prevalence of CPVT is much lower, this could be the main reason that less clinical and functional data have been published in last 5 years.
Long QT syndrome
In our cohort of LQTS, the same three P rare variants remained after five years, making up 12% of the total number of LQTS variants. Of the LP variants, 8 (100%) retained their classification. In a study focused on a cohort of children diagnosed with LQTS, 66.66% of VUS were reclassified as LP/P (Bennett et al., 2019). In contrast, another study reported 14.3% of variants previously classified as LP/P were downgraded to VUS (Westphal et al., 2020). This ambiguity reinforces the need for additional studies focused on a proper reinterpretation, especially if variants were not originally classified following ACMG recommendations (Campuzano, Sarquella-Brugada, Fernandez-Falgueras, et al., 2020, VanDyke et al., 2020). In our cohort, LP/P rare variants remain mainly located in KCNH2, and KCNQ1, consistent with recent critical reappraisals of genes implicated in LQTS (Adler et al., 2020, Giudicessi, Wilde, & Ackerman, 2018). Concerning VUS, we identified two rare variants were reclassified to LB due to new available MAF, supporting periodic checking of available frequency data in order to clarify the role of ambiguous variants.
Clinical translation
The main aim of a comprehensive genetic interpretation is clinical translation. Clinical features and phenotypic context are critical for effectively interpreting the clinical impact of rare variants in ICC, helping to improve the diagnosis, treatment, and family screening of patients who currently receive uncertain genetic test results (Walsh et al., 2021). A recent study focused on the reclassification of VUS in ICC concluded that disease-specific phenotypes significantly increase the accuracy of classification and reinforce the need for clinical data in genetic diagnoses, aiding variant interpretation (Bennett et al., 2019). Hence, updating the classification of a rare variant may have significant clinical impact on patients and their relatives. Therapeutic management can be modified, but emotional and psychological impacts may have lasting effects (Vears, Niemiec, Howard, & Borry, 2018). Another important point is the economic impact of reinterpretation. There is currently no legal duty for laboratories to reanalyze data neither in clinical nor in forensic cases (Giesbertz, van Harten, & Bredenoord, 2019). Hence, the real cost and who pays for the service of reassessing genetic variants over time remains unclear (Vears, Senecal, & Borry, 2020). A reinterpretation can lead to recommendations for preventive measures and health care treatments that were not considered to be necessary until after reanalysis (Pagan et al., 2020). Despite this, there are no current guidelines informing the cost-effectiveness of reanalysis.
Finally, it is important to clarify that new genetic data should be discussed by a group of experts in ICC (Muller, McDonald, Pope, & Cragun, 2020) and, to avoid misinterpretation, an expert cardiologist in genetics should explain to patients what reclassification entails (Bombard et al., 2019, Burns, James, & Ingles, 2018). These expectations should be explicitly delineated as part of the informed consent process before a sample is obtained and reviewed again when disclosing initial results (David et al., 2019). From our point of view, care of ICC families should follow the ethical premise that improving medical care should be the major reason for establishing clinical or research guidelines.
Our study has some limitations. First, variant classification is subject to inherent intra- and inter-laboratory differences in data interpretation.(Amendola et al., 2016) To minimize this in our study, five of the authors (OC, GSB, EA, AF, and RB) performed independent classification following ACMG recommendations, and all authors included in the authorship came to a consensus regarding the final classification decision. Despite this fact, we assume that some variants may be interpreted in a different way in other laboratories, especially concerning VUS variants. Second, concerning genetic diagnosis, it should be noted that patients may carry additional rare variants in ICC-related genes that are currently unknown and hence not included in our gene panel. Third, it is important to consider that our cohort was limited and comprehensive reassessment should be performed in larger ICC cohorts to corroborate periodic reclassification. Despite not concrete time, similar studies also recommed periodic reanalysis. We propose, for the first time, a time frame for genetic reanalysis in ICC. Fourth, we have not assessed the economic cost of a comprehensive reinterpretation and who (government, hospital/health institution, or patient) should assume this cost. This is a controverse point that should be deeply analyzed due to each country has a particular health system. Finally, lack of available data for some rare variants included in our study currently impedes more accurate interpretation, remaining as ambigous.
In summary, a comprehensive genetic interpretation of rare variants associated with inherited channelopathies is warranted because it has practical consequences for patients and their relatives. We found that 18.36% of rare variants changed classification, based on the current ACMG criteria, within five years. These changes were mainly due to new data on the global frequencies of rare variants. However, many rare variants remain of ambiguous significance due to a lack of functional data and, most importantly, conclusive family segregation. We recommend performing reanalysis of rare variants associated with inherited channelopathies at least every five years to incorporate new data concerning pathogenicity. When significant changes in classification occur, the cardiologist should promptly inform interested patients and, if necessary, modify the therapeutic approach.