Role of the proportion of sudden cardiac death to mortality for clinical effectiveness of primary prevention ICDs

This commentary refers to ‘Clinical effectiveness of primary prevention implantable cardioverter-defibrillators: results of the EU-CERT-ICD controlled multicentre cohort study’, by M. Zabel et al., doi: 10.1093/eurheartj/ehaa226. We thank Disertori et al.1 for their Discussion Forum comments regarding our European Heart Journal 20202 publication. Among EU-CERT-ICD patients, implantable cardioverter-defibrillator (ICD) therapy was associated with a significant 27% total mortality reduction. In pre-specified patient subgroups, e.g. patients aged ≥75 years or diabetics, this benefit seemed reduced or lost indicating some heterogeneity of the ICD effect across the study population. Importantly, no harm was seen in any patient groups. Disertori et al. critically comment that the survival benefit found in our study cannot be fully explained by the ICD-related sudden cardiac death (SCD) reduction and might therefore be overestimated. We recognized all-cause mortality as the most reliable primary endpoint. Even in cases of study dropouts, the patients’ vital status can reliably be tracked through local authority registries. In contrast, classification to SCD, non-sudden and non-cardiac mortality requires committee adjudication using circumstantial evidence that might be open to discussion even if collected as carefully as in EU-CERT-ICD. An annualized all-cause mortality of 5.5% among the ICD patients compared with 9.2% among the controls (unadjusted hazard ratio 0.682). The secondary SCD endpoint appeared in 2.6% and 0.5% of control and ICD patients, respectively [unadjusted hazard ratio (HR) 0.158]. Thus, almost 60% of the ICD mortality benefit in our study is explained through SCD reduction. Our SCD to total mortality ratio was similar to randomized ICD studies.3 , 4 Searching for antiarrhythmic ICD effects, the above annualized SCD rate of 2.6% in controls compared favourably with an annualized 2.8% rate of first appropriate shocks in the ICD group (1:1.08). Several patients received more than one appropriate shock, increasing the possible number of SCDs averted. Compared with previous studies, mandatory modern ICD programming likely increased the rate of life-saving shocks in EU-CERT-ICD.2 The proportion of SCD control patients compared with those with appropriate ICD shocks in the ICD group was only 1:2.1 in DEFINITE,4 while improving to 1:1.4 in DANISH.3 Moreover, the conjecture that ICD implantation can save life exclusively by arrhythmia terminating shocks might be too mechanistic. Patients’ perception of ICD-related safety net might potentially enhance their vagal tone through cortex mechanisms with possible benefit to heart failure progression. Clinical monitoring differences between ICD patients and controls might also contribute. In non-randomized controlled studies, potential confounder bias needs to be considered carefully. EU-CERT-ICD enrolled ICD and control groups relatively balanced for many clinical factors. A multivariable model was used for adjustment with sensitivity analyses by several propensity scoring techniques. All methods revealed similar survival benefit between 27% and 32% (HRs between 0.682 and 0.731 on five comparisons).2 In summary, we do not believe that the mortality benefits found in our study are explained by confounding factors and are overestimated. Further research—e.g. the large RESET-CRT and PROFID studies—needs to be completed to balance clinical cofactors for primary prophylactic ICD indications in the guidelines. A careful risk stratification, including advanced electrocardiographic methods5 or late-gadolinium enhancement appears useful. Seattle Proportional Risk Score6 can be recommended to distinguish patients with increased or decreased proportion of arrhythmic death to total mortality. This work was supported by the European Community’s Seventh Framework Programme FP7/2007-2013 (grant agreement no. HEALTH F2-2009-602299) for 5 years (starting 1 October 2013). Conflict of interest: M.Z.: grants from European Union 7th framework program, during the conduct of the study; minor travel grants from Biotronik, outside the submitted work. T.F: grants from European Commission, during the conduct of the study; personal fees from Novartis, Bayer, Janssen, SGS, Roche, Boehringer Ingelheim, Daiichi-Sankyo, Galapagos, Penumbra, Parexel, Vifor, Biosense Webster, CSL Behring, Fresenius Kabi and Coherex Medical, outside the submitted work; H.H.: nothing to disclose. M.M.: grants from European Union 7th framework program, during the conduct of the study. R.W.: grants from European Community’s Seventh Framework Programme FP7/2007-2013 and Postdoctoral Clinical Research FWO-Flanders, during the conduct of the study; grants and other from Medtronic and Biotronik, other from Abbott and Boston Scientific, outside the submitted work.