Additional evidence for management of heart failure - and the use of echocardiographic left ventricular function

Emil Erik Nielsen

Research output: ThesisPh.D. thesis

19 Downloads (Pure)



Heart failure has an estimated prevalence of more than 37.7 million people worldwide in 2010 and is rapidly growing.The growing prevalence of heart failure can be accounted to increased life-expectancy in the general population andincreased prevalence of risk factors such as hypertension, ischaemic heart disease, diabetes mellitus, and metabolicsyndrome. Despite the addition of new drugs, devices and early diagnosis of heart failure, the overall survival rate ofpatients with heart failure has only shown modest improvement. Heart failure has traditionally been categorized according to the left ventricular ejection fraction (LVEF), and the guideline recommended treatment for heart failure havebeen based on this classification. 

The primary choice of the inhibitor of the RAAS has been heavily debated. The first drug inhibiting the RAAS was theangiotensin-converting enzyme inhibitors (ACE-I). Since then, first ARBs, and ARNIs have been introduced. The recommendations the European Society of Cardiology (ESC) and the newest update from the American College of Cardiology(ACC) differs in recommendation of RAAS inhibition.

Objective and methodology

To evaluate whether a simplified echocardiography using the “traditional” measure of ‘’LVEF” is clinically sufficient or amore comprehensive characterization of patients with heart failure using more variables including LV global longitudinalstrain (GLS) might have additional clinical benefits. We planned to evaluate this by performing a systematic review ofnon-randomized studies and a retrospective study that was abandoned due to missing data making the retrospectivestudy inconclusive.

As an alternative we conducted a prospective proof of concept study of which preliminary data is presented in thisthesis. Only baseline data of the preliminary data were available. Therefore, at the time of submission of this thesis itwas not judged to be ready for publication. Submission is planned when follow-up data of both echocardiography aswell as supplementary data are available. Therefore, we ended up with the following two studies:

1. ‘The predictive value of global longitudinal strain in patients with heart failure: a systematic review of predictorstudies with meta-analysis’

2. Preliminary data from ‘The clinical value of a detailed echocardiographic classification of patients with type 2diabetes and cardiovascular disease (proof of concept)’.

In addition, we wanted to evaluate the evidence for the medication used for RAAS inhibition, which is the current standard therapy for patients with heart failure, by conducting three systematic reviews.


Study I – Systematic review of GLS as a prognostic marker

In total 16 trials reported data on GLS as a prognostic factor in patients with heart failure. Meta-analyses showed evidence of an association of GLS as a prognostic factor on mortality in a multivariate analyses (Hazard ratio (HR) 0.75; 95%Confidence interval (CI) 0.70 to 0.81; I2 = 9%). No significant statistical heterogeneity was observed between the includedstudies. In trials reporting HRs for both GLS and LVEF adjusted for the same variables, GLS showed statistical associationon mortality, while LVEF did not.

Study II – Preliminary proof of concept study

A total of 41 participants were included in the DIACOR project, and had an echocardiographic assessment conducted.The mean value of LVEF were 47.3% categorized as borderline reduced, while the mean absolute GLS were 12.4 indicating moderate to severe reduction in GLS. Univariate linear regression and visually assessment of the scatter plot indicated that the severity of LV systolic function was more often abnormal when using GLS as a diagnostic tool comparedto LVEF. 

Study III – Systematic review of ARNI (sacubitril/valsartan) versus control

We included 48 trials assessing sacubitril/valsartan versus ACE-Is/ARBs. Meta-analysis and Trial Sequential analysesshowed evidence of a beneficial effect of sacubitril/valsartan in participants with heart failure with reduced ejectionfraction (HFrEF) when assessing both all-cause mortality (Relative risk (RR), 0.86; 95% CI, 0.79 to 0.94) and serious adverse events (RR, 0.89; 95% CI, 0.86 to 0.94). Subgroup analysis indicated that the results did not differ between theguideline recommended target population and HFrEF participants in general.

Study IV – Systematic review of ARBs compared to placebo

We included 244 trials randomizing 32,104 participants with heart failure to ARBs versus control (placebo or no intervention). When including all trials meta-analysis showed evidence of a beneficial effect of ARBs on all-cause mortalityin patients with HFrEF. Subgroup analyses showed significant difference between English-language trials and Chineselanguage trials. Analyzed separately, meta-analysis of English-language trials did not show evidence of an effect (RR,0.91; 95% CI, 0.79 to 1.04) while Chinese-language trials showed a beneficial effect (RR, 0.35; 95% CI, 0.24 to 0.51). Theexplanation for the observed heterogeneity could not be identified.

Study V – Systematic review of ARBs compared to ACE-I

We included 193 trials randomizing 24,202 participants with heart failure to ARBs versus ACE-Is. Meta-analyses and TrialSequential Analysis showed that we could reject any difference between the compared intervention when assessing allcause mortality (RR, 1.01; 95% CI, 0.94 to 1.08; P=0.82; I2=20%).


Results from our systematic review of non-randomized studies indicates that GLS seems to be a significant prognosticfactor for mortality for patients with heart failure irrespective of classification based on LVEF. 

Preliminary data from our prospective proof of concept study including patients with type 2 diabetes and concomitantcardiovascular disease indicates that GLS showed a worse systolic dysfunction compared to the traditional LVEF. Thismight indicate that if GLS is a better prognostic marker, adding GLS may provide valuable clinical information.

Evidence from our reviews showed that ARBs may be beneficial for patients with HFrEF compared to placebo, but theevidence is uncertain. Similarly, head-to-head trials indicate that ARBs do not seem to be neither inferior nor superiorcompared with ACE-Is. The uncertainty of ARBs strengthens the choice of ACE-I as first choice. Sacubitril/valsartan compared with either ACE-Is or ARBs seem to have a beneficial effect in HFrEF patients, and our results indicate that sacubitril/valsartan might be beneficial in a wider population of heart failure. However, several uncertainties need to beresolved in randomized clinical trials before sacubitril/valsartan can be considered first choice.

Neither ARBs nor sacubitril/valsartan seems to be beneficial in heart failure with preserved ejection fraction (HFpEF)patients, while insufficient evidence is available in heart failure with mildly reduced (mid-range) ejection fraction(HFmrEF) patients. The large heterogeneity of the HFpEF and HFmrEF populations, in addition to the results of our systematic review of GLS, suggest that the potential next approach in order to guide medical therapy in these patientgroups might be to conduct randomized clinical trials using a more comprehensive phenotyping including prognosticmarkers such as GLS.
Original languageEnglish
Awarding Institution
  • University of Southern Denmark
  • Olsen, Michael Hecht, Principal supervisor
  • Jakobsen, Janus Christian, Co-supervisor
  • Raymond, Ilan E., Co-supervisor, External person
  • Steensgaard-Hansen, Frank Victor, Co-supervisor, External person
Publication statusPublished - 29. Aug 2022

Note re. dissertation

Print copy of the full thesis is restricted to reference use in the Library.


Dive into the research topics of 'Additional evidence for management of heart failure - and the use of echocardiographic left ventricular function'. Together they form a unique fingerprint.

Cite this