CRT May Work in Preserved-EF Heart Failure, Hints Swedish Registry

October 15, 2012

October 15, 2012 (Stockholm, Sweden) — From the Swedish Heart Failure Registry come suggestions that cardiac resynchronization therapy (CRT) could potentially improve heart-failure outcomes in a far broader population than are currently eligible for it, according to researchers [1].

They propose, extrapolating from their data, that ventricular dyssynchrony based on the QRS interval may be independent of LVEF, and so CRT could potentially benefit in heart failure with preserved ejection fraction (HF-PEF).

There are observational data suggesting CRT can at least improve symptoms in HF-PEF, but it's a big step beyond data from the major CRT trials and currently accepted indications for CRT, which cover patients with an LVEF no higher than 30% or 35%.

Their proposal depends on QRS prolongation, which on the ECG reflects intraventricular conduction delay, as a surrogate marker of dyssynchrony amenable to CRT. In practice, QRS prolongation is a major determinant of whether a patient with systolic HF should be considered for CRT, but its relationship to mechanical dyssynchrony is generally thought to be weak.

CRT itself wasn't considered in the current analysis, which therefore can't speak directly to the device therapy's effects in their cohort, lead author Dr Lars H Lund (Karolinska Institutet, Stockholm, Sweden) observed for heartwire . But their data do show that QRS prolongation to >120 ms predicts all-cause mortality independently of at least 40 other predictors, including LVEF.

"We can be very sure that it's a [mortality] risk factor and not just a marker of something else. And since it's a risk factor, that suggests it's the QRS prolongation and dyssynchrony itself that lead to adverse outcomes," Lund said. And if that's true, "it suggests that treating the dyssynchrony would reduce those adverse outcomes. So we have a strong rationale for why CRT might be of benefit in preserved ejection fraction, and I think it perfectly sets the stage for such a trial."

Hazard Ratio* (HR) for Death From any Cause Associated With Prolonged QRS as Binary and Continuous Variables

Format for QRS as variable HR (95% CI) p
Binary, >120 ms vs <120 ms 1.11 (1.04–1.18) <0.001
Continuous, per 10-ms increment 1.03 (1.01–1.04) <0.001

*Multivariable analyses adjusted for 40 variables including age, sex, LVEF, NYHA class, type of structural heart disorder, history of cardiac procedures, drug therapy, laboratory markers such as creatinine clearance and natriuretic peptide levels, and HF duration

On the other hand, heart failure of greater severity--that is, higher NYHA class--correlated with greater mortality effect of QRS prolongation in univariate analysis, "but all that disappeared in the multivariable analysis, meaning that NYHA class is only a marker for the risk of QRS prolongation--not a [causative] risk factor," according to Lund.

That "signal" of greater harm from QRS prolongation in milder HF, he said, "goes against conventional wisdom" but suggests "in a highly speculative way" that CRT may benefit milder HF with any degree of QRS prolongation.

The idea goes against evidence that CRT is of little or no value when the QRS is less than 130 ms or 150 ms, even in NYHA class 3 heart failure. But it is also consistent with recent clinical trials like MADIT-CRT and REVERSE, which supported CRT in NYHA class 2, and with recent guidelines expanding CRT in NYHA class 2 to patients with QRS >130 when previously the threshold was QRS >150 ms, Lund said.

"What we show is that milder NYHA class might be associated with higher risk, but that doesn't mean that the benefit of CRT would necessarily be greater. It might be that in milder heart failure, it's just not as easy [for CRT] to remedy the harm from QRS prolongation."

Their analysis included 25 171 patients with heart failure who entered the registry over 11 years at most of the country's hospitals and some of its primary-outpatient clinics; their mean age was 74.6, and 40% were women.

Lund discloses receiving research grants for his institution from Medtronic; disclosures for the coauthors are listed in the paper.

Comments

3090D553-9492-4563-8681-AD288FA52ACE

processing....