This observational retrospective cohort study included 36 patients with definite high-risk ARVC and an ICD followed up in our hospital between January 2000 and January 2016.

Definite ARVC was diagnosed according to the 2010 modification of the Task Force Criteria,4 namely, at least 2 major criteria, 1 major and 2 minor criteria, or 4 minor criteria from different diagnostic categories.

The high-risk criterion was established according to the consensus document of international cardiology societies,12 namely, patients in secondary prevention due to ventricular tachyarrhythmias or previous cardiac arrest or in primary prevention due to severe right and/or left ventricular systolic dysfunction.

Demographic, clinical, anatomical, electrocardiographic, and arrhythmic data were collected, as well as the number and type of diagnostic criteria met by patients at the time of diagnosis. The demographic characteristics were recorded at disease diagnosis. The anatomical characteristics were obtained using Doppler echocardiography or cardiac magnetic resonance imaging.

The first major arrhythmic event was defined as the first sustained ventricular tachycardia, aborted sudden cardiac death, or appropriate therapy in a patient who had not had any previous event (patients with ICDs in primary prevention).

The study meets the ethical standards of the Declaration of Helsinki, 1975.

The genetic analysis was carried out in the index cases through massively parallel next-generation sequencing between December 2013 and June 2016 and included all of the genes associated with the disease or with other cardiomyopathies. The following genes were studied: DSC2, DSG2, DSP, FLNC, JUP, PKP2, PLN, TMEM43, CTNNA3, DES, LMNA, RYR2, TGFB3, TTN, CASQ2, CTNNB1, LDB3, PERP, PKP4, PPP1R13L, and SCN5A. Bidirectional sequencing of the gene fragment (PKP2 in this case) of the previously identified genetic variant was performed in 1 relative of this index case.

The exons and their intron boundaries were captured using a probe library (SureSelect Target Enrichment Kit; Illumina, Agilent Technologies). The sequencing was performed using the Illumina platform (Illumina HiSeq 1500; San Diego, California, United States) with a read length of 2 × 100 bases according to Illumina protocols. The bioinformatic analysis was carried out through different genotyping programs. Frequency-related information from different populations was considered (1000 Genomes Project, Exome Variant Server, Aggregation Consortium). For variant filtering, an internal strategy of variant filtering and prioritization designed by Health in Code was used, with the analysis focusing on the main genes included in the arrhythmogenic cardiomyopathy panel. A team of cardiologists and geneticists evaluated and classified the variants using the variant classification schema of Health in Code adapted from the American College of Medical Genetics.

The variants considered clinically relevant according to patient phenotype were confirmed by Sanger sequencing.

Appropriate events were defined as an ICD intervention involving antitachycardia therapy or as shocks in response to sustained ventricular arrhythmia. Inappropriate events were defined as an ICD intervention with antitachycardia therapy or a shock due to supraventricular tachycardia (atrial fibrillation, atrial or sinus tachycardia) or device malfunction (oversensing or noise).

During follow-up, appropriate ventricular arrhythmic events were recorded (number, time from implantation, cycle length, antitachycardia therapy efficacy, and shocks), as well as inappropriate events (number, time from implantation, and cause), atrial fibrillation or heart failure onset, need for substrate ablation, left ventricular damage, and any deaths or transplants that occurred.

SPSS (version 18.0) was used for the data analysis. Continuous variables are expressed as mean ± standard deviation and categorical variables as absolute values and percentages. The Student t test (for normal distributions) or the Mann-Whitney U test (for nonnormal distributions) was used for the comparison of continuous variables. Categorical variables were compared by means of contingency tables and the chi-square test or Fisher exact test. Arrhythmic event-free time was studied using the Kaplan-Meier method.

Two-sided P values < .05 were considered statistically significant.

Most patients were male (28 patients, 78%) with a mean age at diagnosis of 45 ± 18 years. All had definite ARVC according to the 2010 modification of the Task Force Criteria,4 independently of the genetic result.

The diagnostic criteria and the main clinical characteristics are shown in Table 1, taking into account that no patients underwent endomyocardial biopsy to establish the diagnosis, although it was confirmed a posteriori in 1 patient because heart transplantat was performed.

Regarding disease presentation, most patients were diagnosed on the basis of arrhythmic events: 2 patients had aborted sudden cardiac death and 28 had sustained ventricular tachycardia, with left bundle branch block morphology (20 of the superior axis and 8 of the inferior axis); 4 of these patients were diagnosed after assessment of ventricular extrasystoles, 1 with heart failure during the study and another during the family screening (the single relative of this female index case included in the genetic study, her mother). The first 30 patients were implanted with an ICD for secondary prevention, whereas the remaining 6 received the ICD in primary prevention due to severe ventricular damage (2 biventricular and 4 right ventricular).

The average left ventricular ejection fraction was 58% ± 10%; 6 patients (17%) had left ventricular damage and 2 (6%) had severe ventricular dysfunction (left ventricular ejection fraction < 35%). The average right ventricular ejection fraction was 40% ± 18% on cardiac magnetic resonance imaging, and 18 patients (50%) had severe right ventricular dysfunction (right ventricular ejection fraction < 35%) as measured by echocardiography or magnetic resonance.

Late gadolinium enhancement was found in 10 of the 18 patients who underwent cardiac magnetic resonance imaging: in 7 patients, in the right ventricle; in 2, in both ventricles; and in 1, exclusively in the left ventricle.

Next-generation sequencing detected at least 1 pathogenic or likely pathogenic desmosomal variant in 27 patients (75%): 26 index cases and 1 relative. Of these, plakophilin 2 (PKP2) variants were the most frequent, with 19 variants, followed by 5 desmoglein 2 (DSG2) variants and 3 desmoplakin (DSP) variants.

There were 5 nonsense, 15 frameshift, 1 splice, and 6 missense mutations. Ten novel mutations not previously described were found in 15 patients (71%). One of them (NP_004563.2:p.Glu259Glyfs*77) was detected in 5 index cases of 5 different families in the same province; another (NP_004563.2: p.Gly328Glufs*24) showed cosegregation and, in addition, a family member met the criteria for high-risk disease (right ventricular dysfunction) and was thus included in the genetic study (the single relative of this index case included). Regarding the 8 remaining variants, 49 relatives were studied; 21 were found to be carriers and there was cosegregation in 5 of the variants, but no family member met the criteria for high-risk ARVC.

Figure 1 shows their distribution based on the gene affected and its nature. The characteristics of the variants are detailed in Table 2.

Figure 1.

Distribution of the genetic variants found according to the type of gene affected and its nature. DSG2, desmoglein 2; DSP, desmoplakin; PKP2, plakophilin 2.

(0.05MB).

In addition, desmosomal variants were classified as probably not pathogenic in 3 patients: 2 in plakoglobin (JUP) (NP_068831.1:p.Lys261Trpfs*39 and NP_068831.1:p.Glu146Lys) and another in desmocollin 2 (DSC2) (abnormal splicing intron 9).

In addition to the desmosomal mutations, other variants of nondesmosomal genes were detected in 4 patients. In particular, there were 3 mutations in the gene encoding titin (TTN) and 1 in the ryanodine receptor gene (RYR2), which has not been associated with any clinical characteristics or specific prognoses. Isolated variants were not detected in non-desmosomal genes associated with the disease.

The PKP2 gene mutation was associated with exclusively right ventricular damage (100% right ventricular damage vs 0% left ventricular or biventricular; P < .05). In the 6 patients with left ventricular damage, 2 had a mutation in DSP (67% vs 33%; P = .07) and 1 had a mutation in DSG2; in the remaining 3, the genetic study was negative.

The presence of pathogenic variants, the types of genes affected, or their nature were not associated with an early first arrhythmic event (Figure 2), a distinct clinical presentation, or specific electrocardiographic and anatomical characteristics.

Figure 2.

Kaplan-Meier curve showing the probability of a first major arrhythmic event since birth according to the presence or absence of pathogenic or likely pathogenic variants.

(0.1MB).

During a median follow-up of 5 [1-20] years, 23 patients (64%) had at least 1 appropriate arrhythmic event according to the ICD. The median time to first event was 7 [1-29] months. A total of 321 ventricular arrhythmias were registered and treated, with a mean cycle length of 283.05 ± 35.68ms; 7 patients were admitted for arrhythmic storm and 5 required substrate ablation for repeated ventricular arrhythmia (3 endocardial and 2 epicardial).

All patients were taking beta-blockers, mainly sotalol (21 patients [58%]), and 11 patients (31%) took amiodarone at some point.

The presence or absence of causal desmosomal variants or their nature was not associated with prognostic features determined by the frequency of ventricular arrhythmia occurrence or their earlier development.

Of the 23 patients who had an appropriate arrhythmic event during follow-up, 22 (96%) had received an ICD for secondary prevention vs 1 for primary prevention (P = .02). The event-free survival according to ICD indication is shown in Figure 3.

Figure 3.

Kaplan-Meier curves for appropriate event-free survival from defibrillator implantation according to primary or secondary prevention indication.

(0.08MB).

Four patients had an inappropriate event during follow-up: 3 due to paroxysmal atrial fibrillation and 1 due to sinus tachycardia; 3 patients had device-related complications: 2 infections and 1 lead fracture.

During follow-up, 7 patients had atrial fibrillation and 3 attended the emergency department due to heart failure. There was 1 heart transplant (for refractory heart failure), but there were no deaths.

The present work comprises a retrospective and single-center study, with a small number of patients with high-risk ARVC, and the data were not compared with those of moderate- or low-risk disease. In addition, the study did not include post mortem studies that would most probably have revealed a high arrhythmic risk.

Given the retrospective nature of the study, confounding factors might have affected the results.