New foundational therapy in heart failure with reduced ejection fraction: should we keep following the 2016 European Society of Cardiology Heart Failure Guideline in 2021?

Introduction

The 2016 European Society of Cardiology Heart Failure Guidelines (2016 HF GL) suggest sequential therapy initiation with angiotensinogen converting enzyme inhibitor (ACEi) or angiotensin receptor blocker (ARB), beta-blocker (BB) and mineralocorticoid receptor antagonist (MRA) for patients with heart failure with reduced ejection fraction (HFrEF). Since their publication, major trials established the benefit of sacubitril/valsartan (ARNi) and SGLT2 in HFrEF, and ARNi are suggested to replace ACEi/ARB as first line therapy. So, with HFrEF foundational therapy evolvement, the 2016 HF GL sequential therapy initiation algorithm has been raised into question.

Purpose

To compare in the real-world practice, the effect on all-cause mortality of the simultaneous use of every pharmacological class currently included in the HFrEF foundational therapy with conventional sequential therapy.

Methods

A population of consecutive patients (pts) included in a post-discharge structured follow-up programin in a tertiary center was analyzed. Two groups were defined: 1) patients medicated with all pharmacological classes considered the HFrEF foundational therapy (ARNi, BB, MRA and SGLT2 inhibitor), independently of the dosages – “FT group”; 2) patients medicated with ACEi/ARB, BB and MRA on maximal tolerated doses – “2016 HF GL group”. Pts under other therapeutical combinations were excluded. The study groups were compared with Chi-square and Mann-Whitney tests. Impact on all-cause mortality was established with Kaplan-Meier survival analysis and multivariate Cox regression after adjustment for age, sex and baseline creatinine, NYHA functional class and LVEF.

Results

From 2016 to February 2021, a total of 101 pts with HFrEF were included and followed for 25±16 months. 54 pts were included in the FT group and 47 in the 2016 HF GL. The study population (69.3% males, 64.6±11.4 years) were mainly in NYHA functional class II (48%) and III (48%). The most common HF aethiologies were ischemic heart disease (49.5%) and dilated cardiomyopathy (30.7%), median LVEF was 26% and 22% were under CRT. Baseline characteristics were similar between groups, except for diabetes (more common in FT group, 70 vs 22%, p<0.001). All-cause mortality rate during follow-up was significantly different between two groups: 1.9% in FT group and 17% in the HF GL group (p: 0.047) – Figure 1. The implementation of all foundational therapy classes was an independent protective factor for all-cause mortality (HR 0.41; IQR 0.004–0.468; P: 0.010) in multivariate Cox regression.

Conclusion

This real-world study suggests that conventional sequential therapy suggested by the 2016 HF GL may be less effective on reducing all-cause mortality in HFrEF than simultaneous use of all pharmacological classes that nowadays compose the foundation therapy. These results support the hypothesis of promoting early introduction of all therapy classes followed by a tailored uptitration may be beneficial.

Funding Acknowledgement

Type of funding sources: None. Figure 1

Are we aiming for different metabolic targets in heart failure patients?

Introduction

Metabolic control plays an important role on major cardiovascular events (MACE) prevention. The 2019 ESC guidelines on dyslipidaemia management recommend tighter LDL-cholesterol (LDL-C) control in order to prevent cardiovascular events. However, it is not yet proven that thigh control of dyslipidaemia, glycaemic levels and body mass index (BMI) in Heart Failure (HF) patients (pts) have an impact on prognosis.

Objective

To evaluate the impact of LDL-C, HbA1c and BMI values on HF pts mortality and MACE rates.

Methods

Single centre study that included consecutive pts hospitalized for acute / decompensated chronic HF in a tertiary Hospital between January 2016 to December 2018 and followed for 12 months. The impact of LDL-C, HbA1c and BMI on mortality and MACE was assessed using Cox regression and Kaplan-Meier curve, after adjustment for age, sex, functional class and ejection fraction. A safety cut-off was established when any of these variables was deemed protective using ROC curve analysis.

Results

Two hundred twenty-four patients (71.68±13.45 years, 63.8% males) were included. Eighty-four (37.5%) pts had type 2 diabetes, 39.7% had ischemic heart disease and the median left ventricular ejection fraction was 34% (IQR 25–49.5; 60.3% HFrEF; 13.8% HFmrEF; 22.3% HFpEF). The median BMI was 25.4 kg/m2 (IQR 23.1–30.5), HbA1c, 6.4% (IQR 5.6–6.8) and LDL-C, 89.5 mg/dL (IQR 64–106); 145 (64.7%) pts were medicated with statins. The overall mortality and MACE rates during follow-up were 16.1% and 21.0%, respectively. According to the CV risk classification 39.7% pts were at very high risk and 19.6% pts at high risk. On multivariate analysis HbA1c (HR 1.5 IQR 1.1–1.9; p=0.007) and female sex (HR 9.453 IQR 2.4–37.2; p=0.001) were independent predictors of mortality, whereas LDL-C (OR 1.05 IQR 1.022–1.075; p<0.001) and BMI (OR 1.23 IQR 1.075–1.404; p=0.002) were independent protective factors. LDL-C and BMI had no effect on MACE rates, although HbA1c was an independent predictor of MACE (HR 1.27 IQR 1.03–1.57; p=0.026). For high and very high-risk pts there was still a protective trend on mortality, although non-significant, for higher levels of LDL-C (OR 1.04 IQR 0.99–1.075; P=NS). Protective LDL-C cut-off were estimated for the whole population (LDL-C 88mg/dL; AUC 0.819; sn 56.6%, sp 100%) and for the high and very-high CV risk pts (LDL-C 84mg/dL; AUC 0.815; sn 59.3%; sp 100%). A BMI safety cut-off for mortality of 25.75 kg/m2 was found (AUC 0.627; sn 61.2%; sp 58.3%).

Conclusion

This study supports the theory of the obesity and LDL-C paradox in HF. Lower LDL-C and BMI increased mortality and there is no trade-off effect on MACE rates, supporting the idea that LDL-C and BMI should not be aggressively addressed in HF pts. In our cohort a cut-off level of LDL-C below 88mg/dL is associated with higher mortality. On the other hand, diabetes should be actively treated as HbA1c predicts death and MACE in HF pts.

Funding Acknowledgement

Type of funding source: None

P3808iBox-CRT: Better response, less complicated, equally fast

Introduction

The optimization of the left ventricle (LV) pacing site guided by the electrical delay increases CRT response rate (RR), however it's necessary to develop technology that allows its universal use.

Purpose

The aim is automatically, and operator-independent, access the conduction delay between the right ventricular (RV) stimulus and the LV available veins in order to select the LV pacing site. It is further intended to compare the total procedure and radiation times in relation to an historical control group.

Methods

Prospective, single-center study that included patients undergoing CRT implant according to the current ESC Guidelines. All patients were submitted to a clinical, electrocardiographic and echocardiographic basal evaluation prior to CRT implantation and at 6 months of follow-up.

To evaluate conduction delays between the RV lead and the LV available veins (RV-LV delay), an external interface - intelligent Box for CRT (iBox-CRT) was used. Four measurements in at least two different tributary veins were made. The implant of all the LV leads was guided by the longest measured delay.

A positive response to CRT was defined as an improvement of >10% in left ventricle ejection fraction (LVEF) or a reduction of end-systolic volume (ESV)>15%. The results were compared to a control group (CG) of pts submitted to CRT implantation in the conventional way.

Results

60 patients were included (68.3% males, 38% ischemic, mean age 67.4±10.2 years) and submitted to CRT implant (37 CRT-P; 23 CRT-D). At basal evaluation, LVEF was 28±7%, end-diastolic volume (EDV) was 200±73ml and ESV 145±64ml. CG (n=51) had similar characteristics.

The RR was 85.7%, significantly higher compared to the CG (55.9%, p=0.003). The ESV reduced 38.2±3% in responders vs 5.7±2% in non-responders (NR) (p=0,005), EDV reduced 33.3±16% in responders vs 13.6±10% in NR (p=0.002), the mean LVEF improved 11% in responders vs −1% in NR (p=0.02).

At follow-up, the mean ESV in the study group (SG) was 89±44 ml vs 132±75ml in the CG (p=0.002) and the EDV 136±51 vs 190±78 (p=0.007).

In addition to a much better response rate, the responders in the study group had significantly higher mean LVEF at follow-up (39±11% vs 37±7%, p=0.032).

The mean intra-procedure RV-LV delay was 187±34mseg. In the responder group the baseline delay was usually higher (190±35 msec) vs NR group RV-LV delay (165±23 msec; p=NS).

Compared with CG, the automatic assessment of RV-LV delay with iBox-CRT did not increase fluoroscopy time (15±16min vs 18±16; p=NS) and shortened procedure time (65±34 vs 108±83min, p<0.005).

Conclusions

The iBox-CRT use enabled an automatic and operator independent RV-LV delays measurement, in order to implant the LV lead at the most delayed site. This technique translated into a major increase in CTR response rate, not compromising the procedure duration nor increasing the radiation exposure.