Importance of Diastolic Function for the Prediction of Arrhythmic Death

Authors, Journal, Affiliations, Type, DOI

• Authors: Thomas Pezawas, Achim Leo Burger, Thomas Binder, André Diedrich
• Journal: Circulation: Arrhythmia and Electrophysiology
• Affiliations: Department of Medicine II, Division of Cardiology, Medical University of Vienna, Austria; Vanderbilt Autonomic Dysfunction Center, Nashville, TN
• Type: Prospective, observer-blinded, single-centre observational study (pilot)
• Published: February 2020
• DOI: https://doi.org/10.1161/CIRCEP.119.007757

Overview

This prospective, observer-blinded study enrolled 210 patients with ischaemic cardiomyopathy (n=120), dilated cardiomyopathy (n=60), and normal LVEF controls (n=30) to test whether diastolic dysfunction grading predicts arrhythmic death. Over a mean follow-up of 7.0±2.6 years, grade III diastolic dysfunction (restrictive pattern) was independently associated with a 3.52-fold increased risk for arrhythmic death or resuscitated cardiac arrest (HR 3.52; 95% CI 2.00–6.22; P<0.001). Critically, this association was independent of LVEF — present in both LVEF ≤35% and LVEF >35% subgroups — suggesting diastolic function grading may improve arrhythmic risk stratification beyond the standard LVEF-based ICD threshold. Authors explicitly label this a pilot study requiring prospective randomised validation.

Keywords

Cardiomyopathy, dilated; sudden cardiac death; heart failure; primary prevention; risk assessment; diastolic dysfunction; arrhythmic death; ICD

Key Takeaways

Background

• ICD primary prevention guidelines rely predominantly on LVEF ≤35%, derived from landmark ischaemic trials (MADIT-II, SCD-HeFT).
• LVEF-based stratification has recognised limitations: a substantial absolute number of sudden cardiac deaths occur in patients with LVEF >35%, and many ICD recipients never experience an appropriate shock.
• Data on diastolic dysfunction as a predictor of arrhythmic death were limited to small retrospective studies at the time of this publication — this was the first prospective study addressing this question with long-term follow-up.

Study Design and Population

• Design: Prospective, observer-blinded, single-centre; ethics-approved; test results not disclosed to participants or treating physicians (treatment not guided by study).
• Population: 210 patients; 120 ischaemic CMP, 60 DCM, 30 normal LVEF controls; all had undergone coronary angiography with ventriculography.
• Exclusions: History of sustained VA, permanent AF, ventricular pacing dependence.
• Diastolic grading: Performed blinded to baseline data/outcomes using Doppler echocardiography (mitral inflow, Valsalva, TDI annular e') per contemporary recommendations; graded as Normal, Grade I (impaired relaxation), Grade II (pseudo-normal), Grade III (restrictive).
• Primary endpoint: Time to arrhythmic death (AD) or resuscitated cardiac arrest (RCA).
• Secondary endpoint: Nonarrhythmic death.

Baseline Diastolic Grades

• Normal diastolic function: 23 patients (11.0%)
• Grade I (impaired relaxation): 107 patients (51.0%)
• Grade II (pseudo-normal): 31 patients (14.8%)
• Grade III (restrictive): 49 patients (23.3%)
• Mean LVEF: 38.0±14.7%. Grade III patients had shorter follow-up (3.9±2.4 years) due to higher mortality.

Primary Outcome: Arrhythmic Death / RCA

• Overall follow-up 7.0±2.6 years; 28 arrhythmic deaths and 33 RCAs occurred.
• Cumulative 8-year AD/RCA risk by diastolic grade:
  • Normal: 4.5%
  • Grade I: 21.5%
  • Grade II: 37.2%
  • Grade III: 57.7%
• No patient with normal diastolic function died of arrhythmic death.
• Grade III was the significantly highest-risk group (Kaplan-Meier log-rank P<0.001).

Multivariable Cox Regression (Adjusted Model)

• Confounders included in adjusted model: hypertension (HR 2.82; P=0.093), ICD implantation (HR 3.44; P=0.001), diuretics (HR 1.74; P=0.134), QRS ≥0.12s (HR 2.37; P=0.006).
• Grade III dysfunction: HR 3.52 (95% CI 2.00–6.22; P<0.001) — independently associated with AD/RCA.
• Grade I dysfunction: HR 0.41 (95% CI 0.23–0.71; P=0.002) — significantly protective.
• Grade II: HR 1.03 (P=0.93) — no significant effect.
• Ischaemic CMP (HR 1.00; P=0.994) and DCM (HR 0.91; P=0.764) — aetiology was not a significant confounder.

LVEF Stratification (Key Finding)

• LVEF ≤35% subgroup (n=117): Grade III had significantly higher AD/RCA risk than grades I/II or normal (P=0.001). At 8 years: grade III 60%, grade II 42%, grade I 22%.
• LVEF >35% subgroup (n=93): Grade III had significantly higher AD/RCA risk (P=0.014). At 8 years: grade III 41%, grade I or II 21%.
• Interpretation: Diastolic grading identifies high-risk patients missed by LVEF >35% alone — 41% cumulative arrhythmic event risk in grade III patients who would not qualify for primary prevention ICD under standard criteria.

Nonarrhythmic Mortality

• Grade III also had the highest nonarrhythmic mortality in both LVEF strata (LVEF ≤35%: P=0.009; LVEF >35%: P<0.001).
• In the LVEF >35% group, 78% of grade III patients had died at 8 years.
• Authors note this may be particularly relevant for HFmrEF (LVEF 40–50%), where both systolic and diastolic impairment coexist and treatment options are limited.

Proposed Mechanism

• Diastolic dysfunction reflects myocardial fibrosis and ventricular stiffness.
• Fibrotic tissue creates anatomical substrate for reentry mechanisms and ventricular tachyarrhythmia — explaining the gradient of arrhythmic risk across diastolic grades.

Limitations of the document

• Small sample size (n=210): Authors explicitly describe results as pilot study findings. No firm conclusions can be drawn; large prospective randomised validation is required.
• Single-centre design (Medical University of Vienna) limits generalisability.
• Diastolic grading used pre-2016 ASE criteria — different from the current 2025 ASE two-step algorithm; grade definitions may not map directly onto contemporary classification.
• Non-randomised observational design; residual confounding cannot be excluded despite adjusted model.
• Some RCA events may not have been fatal if ICD-treated — the line between RCA-as-endpoint and ICD-appropriate-therapy introduces potential misclassification.
• Permanent AF and sustained VA prior to enrolment were exclusion criteria — results may not apply to these high-risk subgroups.
• ICD implantation was a significant confounder (HR 3.44) — paradoxically, ICD patients had higher arrhythmic events because ICDs were implanted in higher-risk patients, not because ICDs cause events.

Key Concepts Mentioned

• concepts/LV-Diastolic-Function — diastolic grading methodology; grade III as arrhythmic risk marker
• concepts/VA-Risk-Stratification-DCM — LVEF-independent predictors of arrhythmic death

Key Entities Mentioned

• entities/ICD — primary prevention thresholds; diastolic function as candidate additional criterion
• MADIT-II — ischaemic primary prevention ICD basis; cited as limitation of LVEF-centric approach
• SCD-HeFT — ischaemic/NICM primary prevention ICD basis

Wiki Pages Updated

• wiki/sources/arrhythmia-diastolic-circep-2020.md — created
• wiki/concepts/LV-Diastolic-Function.md — added arrhythmic death risk section and contradiction
• wiki/concepts/VA-Risk-Stratification-DCM.md — added diastolic dysfunction as LVEF-independent predictor
• wiki/entities/ICD.md — added contradiction re: diastolic grading in LVEF >35% subgroup
• wiki/sourceindex.md — new entry added
• wiki/wikiindex.md — updated LV-Diastolic-Function and VA-Risk-Stratification-DCM entries