Defibrillator Implantation in Patients with Nonischemic Systolic Heart Failure (DANISH Trial)

Authors, Journal, Affiliations, Type, DOI

• Lars Køber, Jens Jakob Thune, Jens Cosedis Nielsen, and collaborators
• New England Journal of Medicine, 2016; 375(13):1221–1230
• 5 Danish tertiary centres
• Investigator-initiated, open-label randomised controlled trial
• Funding: Medtronic, St. Jude Medical, TrygFonden, Danish Heart Foundation
• DOI: https://doi.org/10.1056/NEJMoa1608029

Overview

DANISH was a multicentre open-label RCT in 1,116 patients with symptomatic non-ischaemic systolic heart failure (LVEF ≤35%, NYHA II–IV, NT-proBNP >200 ng/L) randomised to ICD or usual care. Over a median follow-up of 67.6 months (longest of the major primary prevention ICD trials), ICD implantation did not reduce all-cause mortality (HR 0.87; P=0.28), despite halving the rate of sudden cardiac death (HR 0.50; P=0.005). The neutral primary endpoint is attributed partly to a high competing mortality from non-CV causes (31%) and partly to the pre-SGLT2i/ARNi era enrolment (2008–2014), which may have underestimated background GDMT-mediated SCD reduction. A significant all-cause mortality benefit was observed only in patients aged <68 years (HR 0.64; P=0.01).

Keywords

Non-ischaemic cardiomyopathy, ICD, primary prevention, sudden cardiac death, heart failure, GDMT, NT-proBNP, CRT

Key Takeaways

Study Design and Population

• n=1,116: 556 ICD arm, 560 control arm; open-label; 5 Danish centres; enrolled 2008–2014
• LVEF ≤35%; NYHA class II–IV; NT-proBNP >200 ng/L; non-ischaemic aetiology confirmed
• 58% of both arms received CRT (CRT-D in ICD arm; CRT-P in control arm)
• Mean age ~63 years; ~70% male; mean LVEF ~25%; median follow-up 67.6 months

Primary Outcome — All-Cause Mortality

• ICD arm: 120 deaths (21.6%); Control arm: 131 deaths (23.4%)
• HR 0.87 (95% CI 0.68–1.12; P=0.28) — NOT significant
• This was the primary pre-specified endpoint; trial was powered for all-cause mortality, not SCD alone

Secondary Outcomes

• Sudden cardiac death (SCD): ICD 24 events vs control 46 events; HR 0.50 (95% CI 0.31–0.82; P=0.005) — SCD halved
• CV death: HR 0.77 (95% CI 0.57–1.05; P=0.10) — not significant
• Non-CV death: 31% of all deaths were non-CV; high competing mortality attenuated the all-cause mortality signal

Age Subgroup Analysis

• Age <68 years (median age cut): HR 0.64 (95% CI 0.45–0.90; P=0.01) — significant all-cause mortality benefit
• Age ≥68 years: HR 1.05 (95% CI 0.68–1.62; P=0.84) — no benefit; competing mortality dominant
• CRT subgroup interaction P=0.73 — benefit consistent regardless of CRT status

Harms

• Device infections (non-CRT ICD): HR 6.35 vs control (P=0.006) — sixfold higher infection risk
• Inappropriate shocks: 5.9% in the ICD arm over follow-up

Context and Limitations

• Enrolled 2008–2014: before SGLT2i and sacubitril-valsartan were standard GDMT
• Neither SGLT2i nor ARNi was used; background therapy was ACEi/ARB + beta-blocker + MRA
• SGLT2i and ARNi independently improve LVEF and may reduce SCD risk, potentially narrowing the residual ICD benefit margin further in contemporary practice
• CRT-P in control arm conferred independent mortality benefit, partly explaining why ICD-on-top-of-CRT added less than expected

Limitations of the document

• Open-label design with potential ascertainment bias for cause of death classification
• 58% CRT rate in both arms limits generalisability to non-CRT NICM populations
• Enrolled before SGLT2i and ARNi became standard; modern quadruple GDMT likely reduces background SCD further
• Age subgroup (age <68) is post-hoc/exploratory; requires prospective validation
• Device infection comparator is ICD vs CRT-P (not ICD vs no device), complicating harm interpretation for non-CRT-eligible patients
• Pre-SGLT2i era reduces direct applicability to contemporary patient populations

Key Concepts Mentioned

• concepts/VA-Risk-Stratification-DCM — LVEF-based ICD threshold inadequate for NICM; age and competing risk matter

Key Entities Mentioned

• entities/ICD — primary prevention NICM; Class IIa ESC reflects DANISH neutral primary endpoint
• entities/HFrEF — GDMT context; DANISH predates modern quadruple therapy

Wiki Pages Updated

• wiki/sources/icd-nicm-danish-nejm-2016.md — created
• wiki/entities/HFrEF.md — ICD section enriched; contradiction updated; source added
• wiki/entities/ICD.md — DANISH caveat enriched; contradiction updated; source added; source_count 2→3
• wiki/concepts/VA-Risk-Stratification-DCM.md — DANISH direct citation added; source_count 5→6
• wiki/sourceindex.md — new entry
• wiki/wikiindex.md — ICD and VA-Risk-Stratification-DCM entries updated