Low Natriuretic Peptide Levels and Outcomes in Patients With Heart Failure and Preserved Ejection Fraction

BACKGROUND Although some patients with heart failure (HF) with mildly reduced/preserved ejection fraction have low natriuretic peptide levels, there are no large-scale systematic studies of how common these individuals are or what happens to them. OBJECTIVES The purpose of this study was to examine the proportion of patients in the I-PRESERVE (Irbesartan in Heart Failure with Preserved Ejection Fraction) trial with an N-terminal pro – B-type natriuretic peptide (NT-proBNP) level < 125 pg/mL, their clinical characteristics, and outcomes. METHODS I-PRESERVE enrolled patients with symptomatic HF and a LVEF $ 45% but who did not have NT-proBNP or body mass index inclusion/exclusion criteria. Baseline NT-proBNP was measured after enrollment but not reported to investigators. The primary outcome in this analysis was the composite of cardiovascular death or HF hospitalization. RESULTS Overall, 808 of 3,480 patients (23.2%) had NT-proBNP < 125 pg/mL. Patients with a low NT-proBNP were younger (68.6 years vs 72.6 years; P < 0.001), were less often men (36.1% vs 40.9%; P ¼ 0.015), and had a higher body mass index (48.4% vs 38.7% obese; P < 0.001) than those with a higher NT-proBNP level. Patients with a low NT-proBNP had less atrial

7][8] Similarly, how the clinical characteristics and outcomes of these individuals compare to patients with a higher NT-proBNP level is poorly studied.The I-PRESERVE (Irbesartan in Heart Failure with Preserved Ejection Fraction; NCT00095238) trial offers a unique opportunity to examine these questions because patients were enrolled in that trial based on a clinical diagnosis of HF, with no natriuretic peptide-based enrolment criterion.Indeed, at the time this trial enrolled patients (2002-2005), routine measurement of natriuretic peptides was uncommon.However, as part of the trial protocol, most patients had a measurement of NT-proBNP in a blood sample collected at baseline (these samples were sent for subsequent assay in a central laboratory and investigators were unaware of the results).Moreover, unlike more recent trials, patients with a high body mass index were not excluded from I-PRESERVE.Therefore, we investigated the proportion of patients with an NT-proBNP <125 pg/mL in the I-PRESERVE trial, their clinical characteristics, and outcomes.

METHODS
0][11] Briefly, 4,128 patients aged $60 years with a left ventricular ejection fraction (LVEF) $45%, signs and symptoms of HF, and corroborating evidence (radiological evidence of pulmonary congestion, electrocardiographic left ventricular hypertrophy or left bundle branch block, or echocardiographic left atrial enlargement of left ventricular hypertrophy) were randomized to 300 mg once daily irbesartan or placebo (Supplemental Table 1).Patients in NYHA functional classes II to IV were eligible, but those in NYHA functional class II were required to have had a hospitalization for HF within the previous 6 months.Patients with a plasma creatinine level of more than 2.5 mg/dL (221 mmol/L) were excluded.
Ethics committees at each participating institute approved the trial, and all patients provided written informed consent.The mean follow-up was of 49.5 months.Irbesartan had no significant effect on the primary composite outcome of death from any cause or hospitalization for a cardiovascular cause, or on any of the prespecified secondary endpoints.
[14][15][16] MEASUREMENT OF NT-proBNP.NT-proBNP was measured during the trial in a central laboratory using a sandwich immunoassay on an Elecsys 2010 platform (Roche Diagnostics).Patients without NT-proBNP at baseline were excluded from the analysis (n ¼ 648).
We divided patients according to their NT-proBNP <125 and $125 pg/mL, which is the cutoff proposed in the guideline and the consensus recommendation for diagnosis of HFpEF. 1,17HOCARDIOGRAPHIC SUBSTUDY.In I-PRESERVE, the echocardiographic substudy was only conducted in patients in normal sinus rhythm to ensure good image quality.9][20][21]

Kondo et al
Low NT-proBNP in HFpEF (MLHFQ) at baseline were compared between patients with NT-proBNP $125 and <125 pg/mL using the Wilcoxon rank sum test, and the score for each of the 21 questions of the MLHFQ were compared using ordinal logistic regression models. 22[25] To evaluate the difference in outcomes according to patients with NT-proBNP $125 pg/mL and <125 pg/mL,  NT-proBNP ¼ N-terminal pro-B-type natriuretic peptide.
MLHFQ ACCORDING TO BASELINE NT-proBNP LEVEL.Total MLHFQ score was similar in patients with a lower compared with higher NT-proBNP (43 [Q1-Q3: 29-58] vs 43 [Q1-Q3: 28-59]) (Table 1).The physical dimensions were also similar in each group.By contrast, across the emotional dimensions, those with a lower NT-proBNP had better scores for some questions (being a burden family or friend, and loss of self-control) but worse for the question about depression (Figure 1).ECHOCARDIOGRAPHIC FINDINGS ACCORDING TO BASELINE NT-proBNP LEVEL.Generally, patients with NT-proBNP <125 pg/mL had less marked echocardiographic abnormalities than those with NT-proBNP $125 pg/mL (Table 2).Left ventricular wall thickness and left ventricular mass were higher than normal in both groups, but the elevation was less in patients with NT-proBNP <125 pg/mL.Despite measurements only being made in patients in sinus rhythm, the left atrial area was increased from normal in both groups but was smaller in those with NT-proBNP <125 pg/mL than those with NT-proBNP $125 pg/mL.Both groups had elevated right ventricular systolic pressure but this pressure was lower in patients with an NT-proBNP <125 pg/mL.
EVENT RATES ACCORDING TO BASELINE NT-proBNP LEVEL.The composite of cardiovascular death or HF hospitalization occurred at a rate of 1.9 (95% CI: 1.5-

2.4) per 100 person-years in patients with an
NT-proBNP <125 pg/mL compared with 8.7 (95% CI: 8.1-9.3) per 100 person-years in those with an NT-proBNP $125 pg/mL (Table 3).The relative risk in patients with NT-proBNP <125 pg/mL was lower for all outcomes examined compared to those with an NT-proBNP $125 pg/mL (Central Illustration, Table 3, Figure 2).For example, after adjustment for other prognostic variables, the HR for the composite of cardiovascular death or HF hospitalization in patients with an NT-proBNP <125 pg/mL was 0.35 (95% CI: 0.27-0.46)compared with those with an NT-proBNP $125 pg/mL.In patients with LVEF $50%, event rates for all prognostic variables became smaller, but a similar trend according to NT-proBNP categories was observed (Supplemental Table 5).1.
Rates of the outcomes of interest in the control-arm group of each trial examined are shown in Figure 3.

DISCUSSION
Among patients enrolled in I-PRESERVE based on a clinical diagnosis of HF, almost one-quarter had an NT-proBNP <125 pg/mL.The clinical characteristics and outcomes of these individuals differed markedly from participants with a higher NT-proBNP.26][27][28] However, there were other notable differences, some of which might have been anticipated (eg, lower frequency of AF and chronic kidney disease) and some which were less predictable (eg, lower prevalence of coronary heart disease and diabetes).Conversely, some differences that may have been expected were not seen.0][31][32] Patients with an NT-proBNP <125 pg/mL were less likely to have been hospitalized recently, had fewer signs of congestion, and had fewer electrocardiographic and echocardiographic abnormalities, but had similarly impaired health status evaluated by MLHFQ.
It is notable that the approximately one-quarter of patients with an NT-proBNP level <125 pg/mL identified in I-PRESERVE were, by protocol, excluded from contemporary HFmrEF/HFpEF trials such as DELIVER (NT-proBNP $300 pg/mL among patients not in AF/ AFL and $600 pg/mL in patients with AF/AFL), and PARAGON-HF (>300 and >900 pg/mL, respectively unless hospitalized for HF within 9 months, in which case >200 and >600 pg/mL, respectively).
Not much is known about these excluded patients.
The best existing information comes from 2 seminal observational cohort studies, which showed that some patients with a definitive diagnosis of HFpEF, defined by the gold standard of elevated resting or exercise-induced pulmonary capillary wedge pressure (PCWP), had low natriuretic peptide levels.
However, these were carried out in single specialized centers (in a single country). 6,7The first was a report   More recently, the Mayo Clinic investigators have used 2-dimensional speckle tracking to show that left ventricular global longitudinal strain, right ventricular free wall strain and left atrial reservoir strain were higher (ie, less abnormal) in HFpEF patients with lower compared with higher natriuretic peptide levels (although left atrial reservoir strain was lower in both groups than in healthy control subjects). 33all outcomes than those with a higher NT-proBNP level.Abbreviation as in Figure 1.

Kondo et al
Low NT-proBNP in HFpEF -2 0 2 4 : --natriuretic peptides demonstrated deterioration in left ventricular global longitudinal strain, right ventricular free wall strain, and left atrial reservoir strain over a median follow-up of 3.1 years, a progression that was not observed in healthy control subjects. 33ese investigators concluded that patients with lower natriuretic peptides represent individuals with an earlier stage of HFpEF than those with higher natriuretic peptide levels, but do experience progression of biventricular and atrial dysfunction over time.We could not examine this in our cohort because we did not have longitudinal echocardiographic or natriuretic peptide measurements.
Previously, when the effect of randomized treatment was examined in I-PRESERVE according to the NT-proBNP level divided at the median (339 pg/mL), irbesartan appeared to be beneficial in patients with a lower NT-proBNP level. 42Because the NT-proBNP <125 pg/mL group (examined in the present study) had many fewer events than in the <339 pg/mL group tween I-PRESERVE and other trials that applied NT-proBNP as an inclusion criteria, we analyzed the PARAGON-HF (Prospective Comparison of ARNI [angiotensin receptor-neprilysin inhibitor] with ARB [angiotensin-receptor blockers] Global Outcomes in HF with Preserved Ejection Fraction; NCT01920711) and DELIVER (Dapagliflozin Evaluation to Improve the Lives of Patients with Preserved Ejection Fraction Peak transmitral early diastolic velocity (E), late diastolic velocity (A), and myocardial long-axis lengthening velocity (E 0 ) were obtained.Right ventricular systolic pressure was estimated using peak tricuspid regurgitant velocity (V) as 4(V 2 ) þ 10 mm Hg.All echocardiograms were analyzed at the echocardiography core laboratory at the University of Maryland.A B B R E V I A T I O N S A N D A C R O N Y M S AF = atrial fibrillation AFL = atrial flutter HF = heart failure HFmrEF = heart failure with mildly reduced ejection fraction HFpEF = heart failure with preserved ejection fraction LVEF = left ventricular ejection fraction NT-proBNP = N-terminal pro-B-type natriuretic peptide PCWP = pulmonary capillary wedge pressure

-2 0 2 4 :
---STUDY OUTCOMES AND DEFINITION.The original primary outcome in I-PRESERVE was the time to a composite of death from any cause or hospitalization for a protocol-specified cardiovascular cause.To compare with other clinical trials, we analyzed the composite of cardiovascular death or HF hospitalization as the primary outcome in the present study.In addition, time to cardiovascular death, HF hospitalization, and all-cause death were evaluated.In each trial, all deaths and HF hospitalizations were adjudicated according to prespecified criteria by members of an independent clinical events committee who were unaware of study group assignments.STATISTICAL ANALYSIS.The distributions of NT-proBNP for patients with and without atrial fibrillation (AF) (including atrial flutter [AFL]) on the baseline electrocardiography for each trial and those of body mass index were evaluated.The association between NT-proBNP and body mass index was analyzed using a regression model with body mass index modelled as a restricted cubic spline with 5 knots.Baseline characteristics and echocardiographic parameters were summarized according to NT-proBNP level ($125 pg/mL vs NT-proBNP <125 pg/ mL) as mean AE SD or median (IQR) for continuous variables and counts and percentages for categorical variables.Differences in baseline characteristics were tested using the Fisher exact or chi-square test for binary or categorical variables, the Student's t-test for normally distributed continuous variables, and the Wilcoxon rank sum test for non-normally distributed continuous variables, respectively.The total score, physical dimension, and emotional dimension of the Minnesota Living with Heart Failure Questionnaire obese [body mass index $30 kg/m 2 ]) compared with those with NT-proBNP $125 pg/mL.The association between body mass index and NT-proBNP at baseline is shown in Supplemental Figure3.Congestion markers, including elevated jugular venous pressure, rales, and pulmonary congestion on the chest x-ray, were less prevalent in patients with NT-proBNP <125 pg/mL.LVEF was slightly higher in patients with a low NT-proBNP.Regarding comorbidities, patients with NT-proBNP <125 pg/mL were less likely to have AF (8.5% vs 35.1%), myocardial infarction, diabetes mellitus, stroke, and chronic obstructive pulmonary disease or asthma.They also had markedly better kidney function and less anemia than patients with an NT-proBNP $125 pg/mL.By CENTRAL ILLUSTRATION Heart Failure With Preserved Ejection Fraction and Low NT-proBNP Levels , LVEF ≥45% NYHA functional class III-IV or II-IV with HF Hospitalization history within 6 months Current signs and symptoms of HF Younger, more often female Higher body mass index Less evidence of congestion Less AF, MI history, DM, COPD, anemia More hypertension Better kidney function Fewer ECG/echocardiographic abnormalities Similarly depressed health status Cardiovascular Death or HF Hospitalization Characteristics of Patients With Low NT-proBNP Levels I-PRESERVE: N = 3,480 Event rate (per 100 person-years) ≥125 pg/mL: 8.7 (95% CI: 8.1-9.3)<125 pg/mL: 1.9 (95% CI: 1.5-2.4)Adjusted HR: 0.35 (95% CI: 0.27-0.46)et al.J Am Coll Cardiol HF. 2024;-(-):---.The left lower panel shows the proportion of patients with an NT-proBNP level <125 and $125 pg/mL.The right lower panel shows the difference in the outcome on cardiovascular death or HF hospitalization according to patients with NT-proBNP <125 and $125 pg/mL.HR was adjusted for treatment assignment, age, sex, region, systolic blood pressure, heart rate, body mass index, NYHA functional class, hospitalization for HF within previous 6 months, LVEF, history of MI, history of hypertension, history of DM, AF or atrial flutter on an electrocardiogram, and estimated glomerular filtration rate.AF ¼ atrial fibrillation; COPD ¼ chronic obstructive pulmonary disease; DM ¼ diabetes mellitus; ECG ¼ electrocardiography; HF ¼ heart failure; I-PRESERVE ¼ Irbesartan in Heart Failure with Preserved Ejection Fraction; LVEF ¼ left ventricular ejection fraction; MI ¼ myocardial infarction; contrast, patients with NT-proBNP <125 pg/mL had a more frequent history of hypertension.Regarding medications, patients with NT-proBNP <125 pg/mL were treated with an angiotensin-converting enzyme inhibitor, beta-blocker, and spironolactone less frequently, but calcium-channel blockers more frequentlypg/mL.Patient characteristics according to NT-proBNP categories among those assigned to irbesartan vs placebo and those aged <70 years vs $70 years were shown in Supplemental Tables 2

FIGURE 1
FIGURE 1 Score for Each Minnesota Living with Heart Failure Questionnaire Question According to Baseline NT-proBNP Level of 159 consecutive patients enrolled in the Northwestern University (Chicago) HFpEF Program with the clinical syndrome of HFpEF and an elevated resting PCWP.Of these, 46 (29%) had B-type natriuretic peptide #100 pg/mL.In a second much larger study of 581 consecutive participants undergoing invasive cardiopulmonary exercise testing for unexplained dyspnea at the Mayo Clinic between February 2006 and March 2018, 3 groups were described: patients with HFpEF and NT-proBNP $125 pg/mL (n ¼ 263, 45%), patients with HFpEF and NT-proBNP <125 pg/mL (n ¼ 157, 27%) and control subjects with normal hemodynamics, ie, PCWP <15 mm Hg at rest and <25 mm Hg on exercise

(n ¼ 161
, 28%).Although these 2 studies were singlecenter studies in the United States, the prevalence of HFpEF with a low natriuretic peptide level was generally similar between these and our study, with a large number of cases from multiple countries.Thus, all 3 studies are complementary and have the same core findings: patients with low natriuretic peptide levels were younger, had better kidney function, were less likely to have coronary artery disease, were much less likely to have AF, and were much more likely to be obese.The proportion of patients with obesity was, however, much greater overall in the 2 U.S. studies than in I-Preserve.All 3 studies showed that patients with lower natriuretic peptide levels have less right ventricular enlargement and, especially, less left atrial enlargement (although left atrial volumes were somewhat greater in I-PRESERVE than in the other 2 studies).

FIGURE 2
FIGURE 2 Kaplan-Meier Curves According to Baseline NT-proBNP Level

(
described in the prior publication), only a trend to better outcomes was apparent in the present study in the irbesartan group.However, this retrospective finding, suggesting an unexpected differential effect of treatment, needs to be treated with great caution, especially as the overall trial was neutral.Nevertheless, the analyses of outcomes in the 2 NT-proBNP groups examined in the present study were adjusted for randomized treatment allocation.STUDY LIMITATIONS.First, patients enrolled in clinical trials are selected according to specific inclusion and exclusion criteria, including requirements other than NT-proBNP and body mass index.Therefore, patients in I-PRESERVE do not reflect the full range of unselected "real world" patients with HFpEF, eg, individuals with a creatinine >2.5 mg/dL.Patients with LVEF in the range of 40% to 44%, included in the current definition of HFmrEF, were not enrolled in I-PRESERVE.Black patients, who are known to have lower levels of natriuretic peptides, were not well represented in I-PRESERVE.In I-PRESERVE, echocardiographic parameters were obtained only for patients in sinus rhythm, which may have resulted in an underestimation of the enlargement of the left atrial area and indices of diastolic dysfunction indexes.We were unable to investigate the detailed biological differences between patients with lower and higher NT-proBNP levels and a recent proteomic investigation has suggested that high NT-proBNP levels in patients with HFpEF are a marker of fibrosis and inflammation.43,44More detailed characterization of potential differences between patients with lower and higher NT-proBNP levels is therefore of interest in future studies, eg, in identifying alternative diagnostic markers and biomarkers improving risk prediction in HFpEF.CONCLUSIONSAlmost one-quarter of patients in I-PRESERVE had a low NT-proBNP level, suggesting that there are many such patients in the "real world" HFmrEF/HFpEF population (and the need for rigorous diagnostic investigation).Although HFmrEF/HFpEF patients with a low NT-proBNP have relatively low rates of hospitalization and death, they have a marked reduction in health status as measured by their MLHFQ.These findings emphasize the important need for treatments that improve symptoms and quality of life in HFmrEF/HFpEF, even among patients with low NT-proBNP.

TABLE 1
Patient Characteristics According to NT-proBNP Categories Continued on the next page

TABLE 1
Values are n (%), mean AE SD, or median (Q1-Q3).a Underweight, body mass index <18.5 kg/m 2 ; normal, body mass index 18.5-25.0kg/m 2 ; overweight, body mass index 25.0-30.0kg/m 2 ; and obese, body mass index $30 kg/m 2 .Heart rate is missing in 3 cases; body mass index in 12 cases; current smoker and NYHA functional class in 1 case; time from diagnosis of heart failure (HF) in 4 cases; LVEF in 3 cases; Minnesota Living with Heart Failure Questionnaire (MLHFQ) score in 791 patients; pulmonary congestion in 118 cases; peripheral edema, third heart sound, elevated jugular venous pressure, and rales in 3 cases; liver enlargement in 10 cases; QRS duration in 211 cases; left bundle branch block and left ventricular hypertrophy on electrocardiography (ECG) in 206 cases; right bundle branch block in 210 cases; estimated glomerular filtration rate (eGFR) and serum creatinine in 23 cases; blood urea nitrogen, albumin, serum sodium in 24 cases; and hemoglobin in 65 cases.ACE ¼ angiotensin-converting enzyme; AF ¼ atrial fibrillation; CABG ¼ coronary artery bypass graft; LVEF ¼ left ventricular ejection fraction; NT-proBNP ¼ N-terminal pro-B-type natriuretic peptide; PCI ¼ percutaneous coronary intervention.

TABLE 2
Echocardiographic Findings According to NT-proBNP Categories Values are mean AE SD.Body surface area was calculated by Du Bois formula: 0.007184 Â weight 0.425 Â height 0.725 .a Full range.b Differences in LV end-diastolic volume index, LV end-systolic volume index, mass index, and LA index.LA ¼ left atrial; LV ¼ left ventricular; MA ¼ mitral annulus; MV ¼ mitral valve; RV ¼ right ventricular; other abbreviation as in Table

TABLE 3
Outcomes According to NT-proBNP Categories Models were adjusted for treatment assignment.bModels were adjusted for treatment assignment, age, sex, region, systolic blood pressure, heart rate, body mass index, NYHA functional class, hospitalization for HF within previous 6 months, left ventricular ejection fraction, history of myocardial infarction, history of hypertension, history of diabetes, atrial fibrillation or atrial flutter in electrocardiography, and estimated glomerular filtration rate.cAllPvalues were <0.001.Ref.¼Reference;other abbreviations as in Table1. a One-quarter of patients with HFmrEF/HFpEF had low NT-proBNP levels, and these patients had lower risks of cardiovascular death and HF hospitalization but had similarly impaired health status compared with those with a higher NT-pro BNP.TRANSLATIONAL OUTLOOK: Treatments that improve symptoms and quality of life are needed even in HFmrEF/HFpEF patients with low NT-proBNP levels who, despite low rates of morbidity and mortality, have substantially impaired health status.38.Snipelisky D, Kelly J, Levine JA, et al.Accelerometer-measured daily activity in heart failure with preserved ejection fraction: clinical correlates and association with standard heart failure severity indices.Circ Heart Fail.2017;10: e003878.39.Redfield MM, Anstrom KJ, Levine JA, et al.Isosorbide mononitrate in heart failure with preserved ejection fraction.N Engl J Med. 2015;373: 2314-2324.40.Redfield MM, Chen HH, Borlaug BA, et al.Effect of phosphodiesterase-5 inhibition on exercise capacity and clinical status in heart failure with preserved ejection fraction: a randomized clinical trial.JAMA.2013;309:1268-1277.41.Majani G, Giardini A, Opasich C, et al.Effect of valsartan on quality of life when added to usual therapy for heart failure: results from the Valsartan Heart Failure Trial.J Card Fail.2005;11:253-259.42.Anand IS, Rector TS, Cleland JG, et al.Prognostic value of baseline plasma amino-terminal pro-brain natriuretic peptide and its interactions with irbesartan treatment effects in patients with heart failure and preserved ejection fraction: findings from the I-PRESERVE trial.Circ Heart Fail.2011;4:569-577.43.Azzo JD, Dib MJ, Zagkos L, et al.Proteomic associations of NT-proBNP (N-terminal pro-Btype natriuretic peptide) in heart failure with preserved ejection fraction.Circ Heart Fail.2024;17:e011146.44.Myhre PL, Omland T, Shah AM.Ongoing enigma of NT-proBNP in HFpEF: insights from proteomics.Circ Heart Fail.2024;17:e011428.KEY WORDS body mass index, heart failure, heart failure with preserved ejection fraction, natriuretic peptide APPENDIX For supplemental figures and tables, please see the online version of this paper.