Discussion.

Our data provide important insights into the prevalence, clinical features and diagnostic evaluation of patients with unexplained ventricular arrhythmia in a real-world analysis. The key finding of this study is that only a minority of patients with UVA underwent a comprehensive assessment with an extended panel of investigations. Furthermore, our data suggest that completing a more comprehensive assessment in patients with UVA may allow for a significant proportion of these patients to be assigned a more specific diagnosis with subsequent targeted treatment.

Prevalence and clinical features of patients with UVA

In the present study, 38% of secondary prevention ICD recipients who were evaluated met our criteria for UVA. This is a higher proportion of UVA than has been demonstrated in existing research. In a retrospective review of 717 survivors of sudden cardiac arrest by Waldmann et al8, the proportion of cases which remained unexplained following assessment of baseline ECG, TTE and coronary arteries was 12.3%. We posit two reasons for this discrepancy. First, our analysis was restricted to ICD recipients under the age of 60. Younger patients are more likely to present at an earlier stage of their cardiac pathology and are thus more likely to have subclinical cardiomyopathies or only subtle baseline ECG changes at their time of presentation. Second, our study only examined patients receiving secondary prevention ICDs, rather than evaluating all-comers with cardiac arrest or ventricular tachyarrhythmia. This excluded patients who presented with ventricular tachyarrhythmias secondary to acute ischaemia or other reversible causes, in whom an ICD would not be implanted. Both features of our analysis may have led to a higher representation of UVA.
Our data suggest that patients with UVA are clinically distinct from their counterparts with ventricular tachyarrhythmias with manifest structural or electrical heart disease. UVA patients in the present study were significantly younger, more often female had a lower burden of traditional cardiovascular risk factors. Our findings are largely in keeping with previous research in this field, which has consistently demonstrated that patients with unexplained or idiopathic ventricular arrhythmia are of younger age at time of presentation8,9. Existing data on sex differences in patients with UVA are more conflicted. While women account for a higher proportion of overall presentations of ventricular arrhythmia with structurally normal hearts10, men account for a higher percentage of patients presenting with unexplained sudden cardiac arrest or death5,6.
A history of a past episode of syncope preceding the acute presentation with ventricular tachyarrhythmia was reported in 23% of patients in the UVA cohort of the present study. This high proportion of prior syncope highlights the propensity for recurrent malignant arrhythmias in this cohort. This is further supported by the fact that patients in the UVA cohort more frequently required device delivered tachy-therapies during follow-up compared to patients with VA of clear aetiology. This higher risk for recurrent arrhythmia was also demonstrated in a previous review of medium-term outcomes in 66 patients with idiopathic ventricular fibrillation, in which recurrent ventricular arrhythmias were seen in 20% of patients6.
Significant differences in the pattern of prescribed therapies were demonstrated between patients with UVA compared to those with VA of clear aetiology. This may be due to the higher proportion of structural heart disease in the VA of clear aetiology group, including coronary artery disease and heart failure with reduced ejection fraction. Both of these conditions represent an indication for beta blockade independent of presentation with ventricular arrhythmia. However, amiodarone was also prescribed at a lower rate to UVA patients. One reason for this may be a reluctance to overprescribe medications to patients with structurally normal heart and no clear diagnosis, particularly in medicines with long-term toxicity such as amiodarone.

Diagnostic evaluation of UVA in current practice

In this single-centre review, 12-lead ECG and transthoracic echocardiogram data was acquired in all patients requiring a secondary prevention ICD. Conversely, coronary evaluation was acquired in most, but not all secondary prevention ICD recipients, and 13% of patients in the UVA cohort did not undergo any form of coronary assessment. Current ACC/AHA provide a Class I recommendation for the use of either CT or invasive coronary angiography in patients with unexplained cardiac arrest7. Potential reasons for patients with UVA not being selected for coronary assessment in the present study include younger age and lack of traditional cardiovascular risk factors. However, non-atherosclerotic coronary disease remains an important cause of ventricular arrhythmia in younger patients, and may be due to anomalous coronary arteries11, coronary embolic events12 or coronary vasospasm13.
In our analysed cohort of 39 patients with UVA, the utilisation of second-line investigations was highly heterogenous. As the statewide CMR quaternary referral centre for Victoria, CMR was the most commonly utilised modality of testing, adopted in 82% of participants. However, other investigations were comparatively underutilised, including genetic testing (26%), flecainide challenge (21%), exercise ECG (21%) and EP study (15.4%). Such inconsistency in the evaluation of UVA has been replicated in existing literature. Waldmann et al demonstrated that in a cohort of 81 cases of unexplained cardiac arrest, while CMR was utilised in 81% of patients, other investigations including ajmaline challenge, EP study and genetic testing were performed in only a minority of cases (43%, 25% and 18% respectively).
Reasons for this variability are uncertain, but several factors are probably involved. First, standardised protocols for the evaluation of UVA are not in place. Such a standardised assessment has been studied and advocated for in the past9, but has not gained traction in current practice. Second, most of these second-line investigations are usually completed in the outpatient setting, where it can be more challenging to organise further investigations and perhaps many patients with UVA may be lost to follow-up, particularly given their younger age. One reason why the uptake of CMR is higher than other modalities may be the impetus to complete this investigation as an inpatient prior to the insertion of an ICD to allow for maximal diagnostic yield.

Diagnostic yield of comprehensive assessment in UVA

In our study, 17 out of 39 patients (43%) of patients with UVA had a cause for ventricular arrhythmia suggested by one of the five second-line investigations which were evaluated. This is in spite of the highly variable nature of the work-up performed, and an even higher proportion of patients may have had a diagnosis confirmed if complete work-up was performed in all patients.
CMR had the highest diagnostic yield of all the investigations studied, suggesting an underlying diagnosis in 8 patients. All diagnoses related to structural heart disease initially not identified on transthoracic echocardiography. This included diagnoses related to enhanced evaluation of right ventricular function (ARVC diagnosed in 2 patients), better visualisation of subtle structural anomalies (mitral annular disjunction diagnosed in 1 patient), evaluation of tissue oedema (acute myocarditis diagnosed in 1 patient) and patterns of late gadolinium enhancement (cardiac sarcoidosis diagnosed in 2 patients, previous myocarditis diagnosed in 1 patient, transmural scar suggestive of prior infarct diagnosed in 1 patient). CMR is well understood to be more sensitive for the diagnosis of cardiac sarcoid14, myocarditis15 and ARVC16 compared to echocardiography, all of which have a strong association with ventricular arrhythmia.
Conversely, exercise ECG had the lowest diagnostic yield in this study, with no additional information gained in 8 patients that underwent this test. Part of this may be explained by the small number of patients who underwent this investigation. While this study alone does not lend support to the utility of exercise ECG in providing diagnostic clarity in patients with UVA, previous research has shown that exercise testing can unmask primary electrical disorders such as catecholaminergic polymorphic VT17 and long QT syndrome18. Furthermore, a negative stress ECG may provide value as a marker of improved prognosis and lower risk of recurrent arrhythmia in patients with previous cardiac arrest19.
Ascertaining the underlying cause for unexplained ventricular arrhythmia is of substantial clinical importance. First, it allows for targeted therapies to be utilised, not only to prevent future arrhythmia but also prevent the development of future structural heart disease. An example of this is prescription of immunosuppressive therapy after diagnosis of active sarcoidosis or myocarditis. Second, it allows for better risk stratification and lifestyle modification advice. For instance, two patients in our cohort were diagnosed with concealed Brugada syndrome with the use of a flecainide challenge, and appropriate advice about avoidance sodium channel blocking drugs and awareness of hyperthermia was provided. Third, it has important implications for family screening, particularly in cases of subclinical non-ischaemic cardiomyopathy (e.g. ARVC) or channelopathies. Fourth, in certain cases, elucidation of the underlying diagnosis may preclude the need for ICD therapy. An example in our cohort was one patient who was found to have a right paraseptal accessory pathway on EP study with malignant anterograde conduction properties. Diagnosis and ablation of this pathway would effectively prevent future pre-excited tachycardia and may have obviated the need for a defibrillator.

Limitations and directions for future research

Our analysis has a number of important limitations. First, this study is limited by the small sample size. This is an inherent limitation of studying ventricular arrhythmia and cardiac arrest in younger patients, which remains a relatively infrequent, albeit important clinical entity.
Second, this was a single-centre retrospective review. The findings relating to the adoption of certain diagnostic tests therefore only reflects the practice at our own institution, and cannot be extrapolated to the practice in the larger Australian or international cardiology community. One important distinction was the high rate of utilisation of CMR in this study, as access to CMR in varies significantly between cardiac centres in Australia. However, it is notable that the adoption rates for CMR, flecainide challenge and genetic testing in our study were similar to the rates described in a previous European retrospective review of patients with unexplained sudden cardiac arrest8. Both this limitation and the small sample size could be mitigated by a multi-centre extension of the current study.
Third, this study did not evaluate the role of early versus delayed diagnosis in patients with UVA. A proportion of the patients in the UVA cohort were diagnosed with an underlying aetiology for ventricular arrhythmia early in their clinical course; for instance, patients who underwent CMR prior to ICD implant while still an inpatient after presenting with cardiac arrest. Such patients would therefore not have had ‘unexplained’ ventricular arrhythmia from this point onwards and during their subsequent follow-up. In future research, this limitation could be mitigated by evaluating patients who had a label of ‘unexplained ventricular arrhythmia’ at their time of hospital discharge. In such an analysis, it would be interesting to measure the time delay between the initial presentation with ventricular arrhythmia and the eventual diagnosis being made.
Finally, our analysis was restricted to patients implanted with a secondary prevention ICD. This was chosen as a practical method to collect data on patients who had suffered from ventricular tachyarrhythmias presenting to our institution. However, this fails to capture patients presenting with ventricular arrhythmia in whom defibrillators are not implanted, and those patients who die from sudden cardiac arrest whose diagnoses may be made on post-mortem autopsy. Future research in this field could instead review registry data for patients with sudden cardiac arrest, to more widely explore the subject of unexplained ventricular arrhythmia.