Author + information
- Cardiology Division, Massachusetts General Hospital and Baim Institute for Clinical Research, Boston, Massachusetts
- ↵∗Address for correspondence:
Dr. James L. Januzzi, Jr., Massachusetts General Hospital, Yawkey 5B, 55 Fruit Street, Boston, Massachusetts 02114.
There are few measures in heart failure (HF) cardiology more debated than the left ventricular ejection fraction (LVEF). On the one hand, there is no other variable more important to define how we treat patients: thanks to its use in clinical trials of HF therapy, LVEF is the sine qua non for guideline-directed medical therapy (GDMT) in heart failure with reduced ejection fraction (HFrEF) (LVEF ≤40%). Together with increased LV volumes, falling EF identifies LV remodeling, a deleterious process that represents a target for the GDMT that unmistakably save lives in those with HFrEF.
On the other hand, a love-hate relationship exists with LVEF: though this measure plays an important role in defining appropriate care for our patients with HF, some have decried its use as the gold standard for evaluating and categorizing patients with HF. Rigid focus on LVEF has, in part, stymied development of therapies for those with heart failure and preserved ejection fraction (HFpEF) (LVEF ≥50%), in part due to heterogeneity of prognostic phenotypes in those affected, and a glaring lack of information exists regarding optimal therapies for those with heart failure with mid-range ejection fraction (HFmrEF) (LVEF 41% to 49%), largely due to previous exclusion of such patients in completed clinical trials of GDMT. Even among those with HFrEF—where LVEF is the trigger for life-saving GDMT—detractors have correctly pointed out that calculations of LVEF are inaccurate, undermined by variables such as LV chamber size, and noted that single measurements of LVEF reveal relatively little regarding prognosis in comparison to other ways of assessing LV function such as deformation/strain imaging or circulating biomarkers such as N-terminal pro–B-type natriuretic peptide (NT-proBNP). These are all fair concerns. Yet, LVEF is the current global currency for assessment of cardiac function, is the foundation for how HF therapies are delivered, and is permanently embedded in clinical practice documents regarding management of HF (1).
Considering that LVEF measurement is vulnerable to inaccuracy, and that this measure might not be as uniquely prognostic as we would like, one might ask how to improve fidelity of its assessment for delivering greater prognostic meaning and provide actionable clinical information. Recent data from several sources unmistakably show that serial assessment of LVEF—rather than a single point in time—provides rather substantial information regarding prognosis (2). In many ways, LVEF behaves like other biomarkers (such as NT-proBNP), where repeated measures are superior to a single point in time; to best understand future prognosis with these markers, it is where one is heading, rather than where one has been. In principle, this is not entirely surprising: LVEF represents a measure of LV remodeling, so its trajectory (either forward, with downward LV function, or reverse with improved LV function) is not only likely to inform outcomes, but also it represents an important biomarker of success or failure of HF therapy, and the subject of ongoing mechanistic studies of newer GDMT agents, including sacubitril/valsartan (3).
In this issue of JACC: Heart Failure, Savarese et al. (4) report prevalence and importance of longitudinal LVEF change among patients in the SwedeHF (Swedish HF Registry). In this registry of 4,942 patients with HFpEF (18%), HFmrEF (19%), and HFrEF (63%), the investigators analyzed clinically acquired serial LVEF data obtained a median of 1 year apart and sought to evaluate how often patients experienced forward remodeling, reverse remodeling, or no remodeling in each LVEF category. They then examined risk for the composite of all-cause mortality/HF hospitalization depending on LVEF trajectory. Lastly, they evaluated for predictors of various LVEF trajectories (upward, downward, no change).
The results of the study are simple, elegantly presented, and of interest—and tell an important story (Figure 1). Approximately 39% of patients with HFpEF transitioned to either HFmrEF (21%) or HFrEF (18%), and a similar proportion of patients with HFmrEF at baseline (37%) remodeled progressively to a follow-up LVEF <40%. Though less frequent, reverse remodeling occurred as well, with 25% of HFmrEF rising to the HFpEF category, and 26% of patients rising from HFrEF to either mid-range (16%) or preserved (10%) categories. Using stable HFrEF as the referent category, any improvement from HFrEF was associated with reduced risk for the composite endpoint with a reduction in risk of 45% (95% confidence interval: 0.47 to 0.64) to 58% (95% confidence interval: 0.47 to 0.64) depending on extent increase of LVEF: the greater the rise in EF, the lower the risk for death or hospitalization. Unsurprisingly, the opposite was also the case; using stable HFpEF as the referent group, progressive remodeling with downward trajectory to HFrEF was deleterious. These results stand well beside other recent data regarding frequency and importance of LVEF trajectory change (well summarized in Supplemental Table 3 of Savarese et al. , which is worth reviewing).
Though a fine reaffirmation of the importance of how LVEF trajectory informs prognosis, more important is how to potentially understand and potentially prevent these changes. The SwedeHF investigators identified important variables associated with either upward or downward LVEF trajectory. These are of relevance to clinicians and tell an important story. Variables independently associated with reverse remodeling include shorter HF duration, nonischemic status, less severe HF symptoms, and the important combination of lower NT-proBNP and better kidney function. In contrast, a falling LVEF was more likely in those with ischemic HF, those with higher NT-proBNP concentrations, and in those with diabetes. Importantly, planned follow-up with a nurse-led clinic was independently associated with rise in LVEF, while no such follow-up was associated with its reduction.
Taken together, these results well describe an important reality: those with worse LV remodeling are more challenging to manage, require more assiduous care, and merit better application of GDMT. Prior studies suggest that when such GDMT is delivered and NT-proBNP is reduced, reverse remodeling occurs with attendant improvement in prognosis (5). If such therapy is not delivered, progressive LV remodeling is accompanied by rising NT-proBNP and worse prognosis (5). It would have been of interest to know how patterns in NT-proBNP reflected patterns in LVEF change and whether greater reduction in NT-proBNP concentrations predicted greater reverse remodeling as has been suggested by recent data. Given that biomarker measurement is easier to obtain than an echocardiogram, such understanding might inform clinicians how to monitor patients using NT-proBNP measurement as a trigger for LVEF reassessment: rising natriuretic peptide concentrations might inform the need for more assiduous follow-up, whereas a stable or falling value might obviate the need for routine LV imaging.
Though a useful study, the work by Savarese et al. (4) has important limitations. In this observational study, all data were obtained as part of standard care; thus, unmeasured variables (including exactly why the echocardiogram was performed) are not considered, and the time frame of the imaging was not standardized. This may be the greatest limitation: 25% of the patients in this study had their 2 echocardiograms only 6 months apart while a similar 25% had a gap of 3 years or longer between echocardiograms. To the extent that LV remodeling may occur relatively quickly, it remains uncertain whether it is sound to consider LVEF changes as well as prognosis with such disparate follow-up periods. In particular, those with longer windows between echocardiograms have survived longer to their second echo than those with a shorter window of follow-up, which influences how to interpret prognostic meaning of any change in EF. Its limitations notwithstanding, the study by Savarese et al. (4) adds to the growing importance of understanding the importance of LV remodeling and prognosis and reiterates the complexity of patients with a worsening LVEF trajectory; such patients particularly deserve careful and deliberate application of GDMT with LVEF as a target for such therapy.
With recent data increasingly focused on the importance of EF trajectory rather than single measurements now informing important clinical meaning, LVEF may have a requiem. Though it has been battered with punches and counterpunches over recent years, like many heavyweight boxing champions over the years, LVEF may be returning to the ring for another round.
↵∗ Editorials published in JACC: Heart Failure reflect the views of the authors and do not necessarily represent the views of JACC: Heart Failure or the American College of Cardiology.
Dr. Januzzi is supported in part by the Hutter Family Professorship; has received research funding from Novartis, Roche Diagnostics, Abbott, Singulex, Siemens, Cleveland Heart Labs, and Prevencio; consulting income from Roche Diagnostics, Abbott, Janssen, and Myokardia; and serves on the endpoint committee/Data and Safety Monitoring Board (DSMB) for Boehringer-Ingelheim, Bayer, AbbVie, and Janssen.
- 2019 American College of Cardiology Foundation
- Yancy C.W.,
- Januzzi J.L. Jr..,
- Allen L.A.,
- et al.
- Lupon J.,
- Gavidia-Bovadilla G.,
- Ferrer E.,
- et al.
- Januzzi J.L.,
- Butler J.,
- Fombu E.,
- et al.
- Savarese G.,
- Vedin O.,
- D'Amario D.,
- et al.
- Daubert M.A.,
- Adams K.,
- Yow E.,
- et al.