Arrhythmic Risk Stratification in Nonischemic Dilated Cardiomyopathy

1. Study hypothesis There is an urgent need for reconsidering risk stratification approaches to DCMP patients, in addition to LVEF in order to identify subgroups at high arrhythmic risk that could benefit from ICD. This task could be achieved either by imaging and/or noninvasive, ECG-related, risk stratification techniques. In this direction, the investigators recently introduced a two-step multifactorial, noninvasive ECG findings leading to programmed ventricular stimulation (PVS), electrophysiology study (EP)-inclusive approach, to modify and protect the truly high risk post-myocardial infarction (MI) CAD patients with a LVEF≥40%. Based on the fact that those DCMP patients inducible on PVS into sustained ventricular tachycardia/fibrillation (VT/Vf) are the ones most likely to receive appropriate ICD therapy regardless of the degree of left ventricular contractile dysfunction or/and the presence of complex ventricular ectopy, the investigators now propose a similar two-step, multifactorial, noninvasive, EP-inclusive approach among the DCMP population with either relatively preserved (35%<LVEF≤50% - Group A) or reduced (LVEF≤35% - Group B) ventricular systolic function. The kernel of this approach lies in the premise that NIRFs detect arrhythmogenic potential and invasive PVS ascertains whether it translates to clinical arrhythmogenesis, combining the formers' sensitivity with the latter's specificity. It is hypothesized that this approach will lead to i. Detection of a high risk subgroup among DCMP patients with 35%<LVEF≤50% that would benefit from an ICD in terms of total mortality (given that their non-SCD mortality component is low as they are at early heart failure stages with relatively preserved systolic function) ii. Stratification of the cohort with LVEF≤35% into clearly discernible arrhythmic risk level groups, with patients in the lowest one having a total mortality mostly due to heart-failure related death and potential ICD advantages negated by potential complications.
2. Population and methods ReCONSIDER is a multicenter, prospective observational trial, aiming to enroll DCMP patients in 17 major electrophysiology centers in Greece. Reimbursement for all devices will be provided by the Hellenic State. The study is under the auspices of the Hellenic Cardiology Society, in whose secure servers the online patient database will be maintained. Inclusion and exclusion criteria are described in the relevant sections.
3. Endpoints i. Primary endpoint: Major arrhythmic event (MAE) occurrence (ICD activation ± sudden cardiac death ± sustained ventricular tachycardia/ventricular fibrillation) ii. Secondary endpoints: Mortality (all-cause and heart failure-related) - Heart failure-related hospitalization - Device related complications (including pocket and lead-related infections, as well as inappropriate therapies) B. Stratification and treatment i. NIRFs

Presence of the following NIRFs will be evaluated:

1. History of syncope and/or presyncope in the absence of prodromal symptoms typical of vagally-mediated events
2. LV end diastolic diameter >60mm on echocardiography
3. LGE presence (qualitative approach - dichotomous value)
4. >30 premature ventricular complexes/hour on 24-hour electrocardiography
5. Presence of non-sustained ventricular tachycardia on 24-hour electrocardiography
6. 2/3 positive criteria for late potentials, either conventional or modified43
7. QTc derived from 24-hour electrocardiography >440ms (men) or >450ms (women)44 according to the Fridericia formula from a signal recorded in 3 pseudo-orthogonal leads
8. Ambulatory T-wave alternans ≥65μV in 2 Holter channels20, 45
9. Standard deviation of normal RR intervals ≤75ms on ambulatory electrocardiography
10. Deceleration capacity ≤4.5ms and heart rate turbulence onset ≥0% and heart rate turbulence slope ≤2.5ms46 ii. Cardiac MRI protocol Cardiac MRI protocol will entail the following: All patients recruited will undergo cMRI in a 1.5 Tesla system. All sites of cMRI fulfil the requirements of expertise and equipment to be involved in the trial. Each participating centre has approval by their respective Ethics Committee. Analysis will be performed with dedicated computer software.

Measurements will consist of the end-diastolic and end-systolic right and left ventricular volumes, the ejection fraction of both ventricles, and will be provided by analysing cine-sequences performed on both ventricles. Endocardium and epicardium will be contoured both in end-diastole and end-systole to define end-diastolic and end-systolic volumes and subsequently LV stroke volume, ejection fraction, and myocardial mass. All parameters will be indexed to body surface area.

Additionally, cardiac MRI studies will include late images after gadolinium infusion, using inversion recovery sequence, with late gadolinium-enhanced area expressed as percentage of the left ventricular mass. Presence of LGE in any extent will constitute a positive study in terms of NIRF presence. LGE localization and pattern will also be recorded for further analysis.

iii. Programmed ventricular stimulation protocol PVS protocol will consist of up to 3 extrastimuli being introduced from 2 different right ventricular sites, preferably the apex and outflow tract. These extrastimuli will follow a drive train consisting of 6 stimuli at 2 cycle lengths (550msec and 400msec) at the first site and at a single cycle length (400msec) at the second site. Coupling intervals between extrastimuli will be shortened by 10msec intervals until refractoriness or an interval of 200msec is reached - whichever occurs first. Additionally, once the procedure is completed,β-agonist infusion (isoproterenol) will commence at rates 1-4μg/min. Once baseline heart rate increases to >120bpm the protocol will be repeated, using 3 extrastimuli from a single site and with a driving train cycle length of 400msec. Patients will be considered inducible and the protocol terminated prematurely if monomorphic ventricular tachycardia (VT - either sustained - >30sec in duration - or requiring termination due to hemodynamic instability) or ventricular fibrillation (Vf) is induced at any stage.

iv. Patient Groups

Based on the above, 6 patient subgroups will be available for comparison:

1. Group A patients - 35%<LVEF≤50% Group 1-A: no NIRFs present - no PVS performed Group 2-A: at least one NIRF present - not inducible (VT/Vf) upon PVS. Group 3-A: at least one NIRF present - inducible upon PVS
2. Group B patients - LVEF≤35% Group 1-B: no NIRFs present - not inducible upon PVS Group 2-B: at least one NIRF present - not inducible upon PVS Group 3-B: inducible upon PVS v. Device allocation and programming All Group B patients will receive an ICD/CRT-D device in accordance with current international guidelines4, All Group 3-A patients (inducible VT/Vf upon PVS) will receive an ICD. Groups 1-A and 2-A (low and intermediate risk, respectively) patients will be observed without an ICD device. Device programming will be uniform and will comprise of 2 detection zones with long detection intervals to both reduce inappropriate shocks and increase likelihood that the events correspond to aborted SCD, 49. More specifically, the same approach as in the PRESERVE EF trial will be implemented, with ventricular tachycardia therapy cycle length to be set to 270-330msec and number of intervals to detect to 32. Ventricular fibrillation therapy cycle length will be set to <270msec and number of intervals to detect to 18 out of 24. In devices with time programming, the same cycle lengths will be used but intervals will be set to 7sec for CLs in the ventricular tachycardia range and to 2.5sec for those in the ventricular fibrillation range. Ventricular tachycardia therapy will consist of several attempts of antitachycardia pacing, followed by cardioversion attempts at progressively increasing energy. High-energy shocks will be administered to terminate ventricular fibrillation. vi. Follow up and statistics Based on previous studies (follow up of 46.9 months in the study by Gatzoulis et al and 55.2 months in the one by Halliday et al), a 4-year or 48-month follow up appears a reasonable choice. Patients will be assessed at enrolment, at 30 days and then at every 180 days, until censoring or follow up completion. Biventricular pacing percentage will be assessed in all CRT-device bearers (target >95%). Extra visits will be scheduled in cases of device activation and clinical events.
3. Power analysis i. Assumptions Thus, 60% of Group A patients are expected to have no NIRFs detected (Group 1-A), 30% to have at least one NIRF present yet be noninducible upon PVS (Group 2-A) and 10% to both have NIRF(s) present and be inducible. On the other hand, 10% of Group B patients are expected to both have no NIRFs and be noninducible (Group 1-B), 60% to have at least 1 NIRF while being noninducible (Group 2-B) and 30% to have both NIRF(s) present and be inducible (Group 3-B).

The investigators hypothesize a corresponding incidence of MAEs ranging from as high as 50% (Group 3-B, annual incidence 12.5%), down to 25% (Group 3-A, annual incidence 8.3% - similar to the 8.2% annual rate observed in the high risk cohort of the PRESERVE EF study), 10% (Group 2-B - annual incidence 2.5%) and even 0% (Groups 1-A, and 1-B), in the high, intermediate, and low risk groups, respectively.

ii. Sample size calculation Based on the above assumptions, power analysis suggests that, in order to achieve 80% power (β=0.2) at a significance level (α) of 0.05, with an allocation ratio [high risk subgroup to all group patients, i.e. subgroup 3:(1+2+3) of 0.1 for Group A patients (35%<LVEF≤50%)], would require, after adjusting for projected 10% losses to follow up and for an interim analysis upon occurrence of half of the anticipated major arrhythmic events (MAEs), enrolment of 450 Group A patients in order to demonstrate the correctness of the study hypothesis for the usefulness of the two-step multifactorial noninvasive EP-inclusive risk stratification approach among the DCMP population with relatively preserved LVEF in identifying those at high risk for MAEs, after a follow up of 4 years.

Regarding Group B, based on encouraging findings from previous similar studies, a superiority approach will be pursued, with high statistical significance and statistical power levels (0.001 and 0.05, respectively), to ensure that there is a significant advantage regarding MAE-free survival for the low and intermediate risk 1-B and 2-B versus the high risk 3-B subgroup. Thus, given competing mortality and device implantation risks, withholding ICD therapy based on the two-step risk stratification approach should not affect total survival. This leads [allocation ratio of 0.3 - 3-B/(1-B+2-B)] to a required sample size of 225 patients, followed up over a 48-month period, again after adjusting for projected losses and an interim analysis.

1. Discussion The proposed tiered, multifactorial, PVS-inclusive two-step approach aspires to combine the advantages of both NIRF detection and PVS inducibility with study design allowing for crucial comparison of prognostic accuracy of various NIRFs and between NIRFs and the combinational algorithm. Furthermore, the ReCONSIDER approach is potentially superior to the cMRI alternative currently under intense focus given that the potential electrical instability related to the abnormal substrate will also be taken into account (assessment for late potential, ventricular extrasystoles, ventricular nonsustained tachycardia presence) with more facets of arrhythmogenesis also considered (repolarization abnormalities, sympathetic overdrive), not detectable by LGE.