論文ID: CR-25-0102
Background: The pathophysiological condition between heart failure (HF) with preserved left ventricular ejection fraction (LVEF; HFpEF) and non-HFpEF is different. To elucidate the prognostic value of monocytes, as representatives of the innate immune system, we examined the association between peripheral monocyte counts and future HF-related events in patients with HF.
Methods and Results: A total of 678 patients with HF referred for hospitalization was enrolled. These patients were categorized into 2 groups according to LVEF: HFpEF, and non-HFpEF. Based on the median monocyte values, we then defined the high monocyte group as having peripheral monocyte counts ≥363/mm3 in patients with non-HFpEF, and as peripheral monocyte counts ≥322/mm3 in patients with HFpEF. There were 200 patients with non-HFpEF and 478 with HFpEF. Based on receiver operating characteristic analysis, patients with non-HFpEF who were in the high peripheral monocyte group had a significantly higher risk of HF-related events compared with those in the low peripheral monocyte group. In contrast, the high and low peripheral monocyte groups for patients with HFpEF had no significant difference in HF-related events. Multivariate Cox hazard analysis identified high peripheral monocyte counts as an independent and significant predictor of future HF-related events only in patients with non-HFpEF.
Conclusions: High peripheral monocyte count was an independent and incremental predictor of HF-related events in non-HFpEF, rather than in patients with HFpEF.
Chronic heart failure (HF) is now the leading cause of death, especially in Japan, and its prevalence is associated with the aging population.1–3 Given that HF-related events are the primary cause of death in patients with chronic HF, risk stratification for future HF-related events in these patients can provide valuable information in the clinical settings. The latest guidelines in various countries suggest that patients with HF should be categorized as HF with reduced ejection fraction (EF; HFrEF; EF ≤40%), HF with mildly reduced EF (HFmrEF; EF 41–49%), and HF with preserved ejection fraction (HFpEF; EF ≥50%).4 It is well known that patients with HFpEF have different underlying causes, demographics, comorbidities, and responses to treatments compared with those with non-HFpEF. However, the differences in pathophysiological mechanisms between non-HFpEF and HFpEF remain unclear.
There is growing evidence to support a pivotal pathogenic role of the immune system and inflammation in the pathogenesis of HF. Inflammation, characterized by early leukocyte recruitment, is associated with vascular endothelial dysfunction and atherosclerosis.5 Previous epidemiologic studies have reported that increased leukocyte counts is a strong and independent risk factor for the progression of atherosclerosis in cardiovascular diseases.6 We have previously reported that high monocyte counts was an independent and incremental predictor of cardiovascular events in patients with coronary artery disease (CAD).7 Furthermore, previous studies also reported that inflammation plays a role in the pathophysiology of HF,8 and increased monocyte levels, reflecting systemic inflammation, are associated with endothelial dysfunction and cardiac fibrosis in HFpEF patients.9 However, few studies have evaluated the prognostic role of monocytes in patients with HF, and no study has examined the prognostic value of monocytes in patients with HF classified according to left ventricular ejection fraction (LVEF). To elucidate the prognostic value of monocytes in HF, we investigated the association of peripheral monocyte counts in patients with HFpEF and non-HFpEF with their future HF-related events in the present study.
All procedures were conducted in accordance with the Declaration of Helsinki and its amendments. The study protocol was approved by the institutional review board of Kumamoto University (approval no. Senshin 2225). This study is registered with the University Hospital Medical Information Network Clinical Trials Registry (UMIN000038771).
Study Patients and ProtocolPatients with HF who were referred for hospitalization at Kumamoto University Hospital between 2006 and 2017 were registered. Compensated patients with HF were registered in this study. Patients were excluded for the following reasons: severe valvular disease, active infective diseases (excluded based on clinical evaluations), which included symptoms indicative of infection (fever, elevated white blood cell counts) and the presence of diagnostic imaging or laboratory findings supporting an active infection, and the use of antibiotics for treating active infections at the time of enrollment), history of malignancy, and end-stage renal disease (estimated glomerular filtration rate [eGFR] <15 mL/min/1.73 m2). It total, 678 patients with HF were enrolled in the present study and were followed prospectively until 2020 or until the occurrence of a HF-related event.
Definition of HFpEFHFpEF is defined as HF with LVEF ≥50%; in contrast, non-HFpEF is defined as HF with LVEF <50%, according to the latest guidelines in various countries, including Japan.
Definition of CADCAD was defined as a history of angina or myocardial ischemia (detected by stress tests) coupled with coronary stenosis of >50% of the vessel diameter detected using coronary angiography or computed tomography coronary angiography scan. Also, CAD was defined as a history of myocardial infarction (MI), percutaneous coronary intervention, or coronary artery bypass grafting.
Follow-up and HF-Related EventsPatients were followed up prospectively at our outpatient clinics until August 2020 or until a HF-related event occurred. Hospitalization for HF decompensation was defined as patients admitted with symptoms typical of HF and objective signs of worsening HF requiring intravenous drug administration.
Echocardiography and Blood SamplingEchocardiography was performed on stable patients by experienced cardiac sonographers without knowledge of the study data. LVEF, the ratio of early transmitral flow velocity to early diastolic mitral annular velocity determined using tissue Doppler (E/e′), and the LV mass index were measured by echocardiography (Vivid 7; GE-Vingmed Ultrasound, Horten, Norway; and Aplio XG; Toshiba, Tokyo, Japan), as reported previously.10 The LV stroke volume (SV) was calculated as the product of the LV outflow tract and the integral of the velocity time. Thus, the SV index (SVI) was defined as SV/body surface area. The Doppler-derived pulmonary artery systolic pressure (PAP) was estimated from the tricuspid regurgitation pressure gradient (TRPG).11 Echocardiography on patients in a stable condition and receiving optimal therapy for HF was performed by 7 experienced cardiac sonographers with no knowledge of the study data. The reproducibility and repeatability of the echo parameters were confirmed by 2 different experienced sonographers.
Both compensated HF patients and acute decompensated HF patients were registered in the present study. At the hospital discharge point when patients were in a compensated condition after implementation of medical therapy for HF, venous samples were collected in a stable, fasting state in the early morning to measure the blood monocyte counts, and plasma biochemical parameters included creatinine, sodium, hemoglobin, and B-type natriuretic peptide (BNP) levels. The blood samples were stored frozen at −80℃ until their analysis.
Statistical AnalysisSPSS software (v. 17.0; SPSS Inc., Chicago, IL, USA) was used for the statistical analysis. Non-normally distributed data were expressed as median (interquartile range). A P value of 0.05 was considered to indicate statistical significance. Univariate and multivariate Cox proportional hazard models were used to estimate hazard ratios (HR) and 95% confidence intervals (CIs) for cardiovascular events in patients with HF after adjustment for other confounders. Kaplan-Meier analysis was used to determine the cumulative incidence of cardiovascular events, and the log-rank test was used to compare the incidence of cardiovascular events between groups. Because BNP was not normally distributed, we calculated the natural logarithmic transformed BNP (Ln-BNP) for regression analysis. Receiver operating characteristic (ROC) curves were constructed for peripheral monocyte counts and Ln-BNP to predict the occurrence of HF-related events. The area under the curve (AUC), sensitivity, and specificity were calculated to predict the ability of peripheral monocyte counts and Ln-BNP to detect patients with HF-related events, with an AUC value of 0.50 indicating no accuracy, and a value of 1.00 indicating maximal accuracy.12 The incremental effect of adding peripheral monocyte counts to Ln-BNP to predict future HF-related events was evaluated by comparing AUC between Ln-BNP only, peripheral monocyte counts only, and peripheral monocyte counts plus Ln-BNP, according to the method of DeLong et al.13 We also assessed the incremental effects of adding peripheral monocyte counts to Ln-BNP to predict HF-related events using the net reclassification index (NRI). The categorial NRI was described as previously reported.14,15 Decision curve analysis (DCA) was used to quantify the clinical usefulness of the prediction models.16
We first categorized patients with HF into 2 types according to LVEF: non-HFpEF and HFpEF. Table 1 shows the baseline characteristics of patients in the non-HFpEF (n=200) and HFpEF (n=478) groups. The prevalence of hypertension, dyslipidemia, and the use of calcium channel blockers was significantly higher in patients with HFpEF (P=0.04, P<0.001, and P<0.001, respectively; Table 1). In contrast, the prevalence of men, diabetes, the usage of angiotensin-converting enzyme inhibitor (ACE-I), angiotensin II receptor blocker (ARB), β-blocker, and diuretics were significantly higher in patients with non-HFpEF (P<0.001, P=0.005, P<0.001, P<0.001, and P<0.001, respectively; Table 1). Furthermore, patients with non-HFpEF had a lower eGFR (58.0±19.3 vs. 62.3±19.8; P=0.009; Table 1), higher BNP (175.3±295.4 vs. 62.3±19.8; P<0.001; Table 1) and peripheral monocyte counts (384±158 vs. 341±138; P=0.001; Table 1) than patients with HFpEF. Echocardiography parameters, such as TRPG and PAP, were not statistically different between these 2 groups.
Baseline Characteristics of Patients With HFpEF Compared With Non-HFpEF Patients
| All patients (n=678) |
HFpEF patients (n=478) |
Non-HFpEF patients (n=200) |
P value | |
|---|---|---|---|---|
| Age (years) | 70.3±10.4 | 71.5±9.3 | 67.3±12.3 | <0.001 |
| Sex, male (%) | 413 (60.9) | 262 (54.9) | 151 (75.3) | <0.001 |
| SBP (mmHg) | 128.0±21.8 | 130.6±21.1 | 121.6±22.3 | <0.001 |
| DBP (mmHg) | 71.3±13.9 | 71.2±12.9 | 71.4±15.8 | 0.8 |
| BMI (kg/m2) | 23.9±3.7 | 24.1±3.6 | 23.4±4.0 | 0.017 |
| Hypertension (%) | 508 (74.9) | 375 (78.4) | 133 (66.7) | 0.004 |
| Diabetes (%) | 236 (34.8) | 149 (31.2) | 87 (43.4) | 0.005 |
| Dyslipidemia (%) | 497 (73.3) | 373 (78.2) | 126 (63.1) | <0.001 |
| Atrial fibrillation (%) | 193 (28.5) | 137 (28.7) | 57 (28.3) | 0.8 |
| ACE-I or ARB (%) | 516 (76.1) | 343 (71.8) | 172 (85.9) | <0.001 |
| CCB (%) | 361 (53.2) | 278 (58.2) | 65 (32.3) | <0.001 |
| β-blocker (%) | 488 (72) | 209 (43.7) | 179 (89.4) | <0.001 |
| Diuretics (%) | 238 (35.1) | 115 (24.1) | 132 (66.2) | <0.001 |
| CAD (%) | 351 (51.8) | 253 (53) | 100 (50) | 0.3 |
| NYHA III or IV (%) | 131 (19.3) | 80 (16.8) | 52 (25.8) | 0.014 |
| eGFR (mL/min/1.73 m2) | 61.3±19.4 | 62.3±19.8 | 58.0±19.3 | 0.009 |
| Hemoglobin (g/dL) | 13.0±2.7 | 12.7±1.8 | 13.4±4.1 | 0.017 |
| Peripheral monocyte counts (n/mm3) | 353±145 | 341±138 | 384±158 | 0.001 |
| hs-CRP (ng/dL) | 0.39±1.62 | 0.41±1.8 | 0.36±0.95 | 0.7 |
| BNP (pg/mL) | 204.3±321.8 | 175.3±295.4 | 284.3±380.8 | <0.001 |
| LVEF (%) | 55.9±12.4 | 62.7±5.8 | 39.6±8.0 | <0.001 |
| LAD (mm) | 39.8±7.1 | 39.6±7.0 | 40.4±7.3 | 0.1 |
| E/e′ | 17.0±6.0 | 17.6±5.1 | 15.7±7.6 | 0.001 |
| SVI (mL/m2) | 37.9±10.4 | 40.2±9.8 | 32.8±10 | <0.001 |
| TR-PG (mmHg) | 25.4±9.1 | 25.2±7.9 | 25.6±11.1 | 0.7 |
| PAP (mmHg) | 31.6±10.2 | 31.5±9.0 | 31.8±12.4 | 0.8 |
| CO (L/min) | 4.0±1.2 | 4.3±1.2 | 3.6±1.0 | <0.001 |
Data are presented as mean±SD, and n (%). ACE-I, angiotensin-converting enzyme inhibitor; ARB, angiotensin II receptor blocker; BMI, body mass index; BNP, B-type natriuretic peptide; CAD, coronary artery disease; CCB, calcium channel blocker; CO, cardiac output; DBP, diastolic blood pressure; eGFR, estimated glomerular filtration rate; HFpEF, heart failure with preserved left ventricular ejection fraction; hs-CRP, high-sensitivity C-reactive protein; LAD, left atrium diameter; LVEF, left ventricular ejection fraction; NYHA, New York Heart Association; PAP, pulmonary artery systolic pressure; SBP, systolic blood pressure; SVI, stroke volume index; TR-PG, tricuspid regurgitation pressure gradient.
Follow up of Non-HFpEF and HFpEF Patients
Follow-up data on HF-related events were available for 200 patients with non-HFpEF and patients with HFpEF. During the follow-up period (11–1,500 days; mean 750 days), 29 (14.6%) HF-related events were recorded (Supplementary Table 1). We defined the high monocyte group as peripheral monocyte counts ≥363/mm3 in patients with non-HFpEF and as peripheral monocyte counts ≥322/mm3 in patients with HFpEF based on the median values. Kaplan-Meier analysis revealed that patients with non-HFpEF and high peripheral monocyte group (≥363/mm3) had a significantly higher risk of HF-related events (log-rank test, P=0.003; Figure 1A). In contrast, the high and low peripheral monocyte groups in patients with HFpEF had no significant difference in HF-related events (log-rank test, P=0.35; Figure 1B). The interaction analysis showed that the P-interaction value was 0.249, indicating a non-significant difference in the relationship between monocyte count and HF-related events between the HFpEF and non-HFpEF groups.
Kaplan-Meier analysis of HF-related events for high and low peripheral monocyte groups in non- HFpEF patients (A), and HFpEF patients (B). HFpEF, heart failure with preserved left ventricular ejection fraction.
Baseline Clinical Characteristics of Non-HFpEF Patients in the High and Low Peripheral Monocyte Groups
We further categorized patients with non-HFpEF in this study into 2 groups (high peripheral monocyte group: monocyte count ≥363/mm3; or low monocyte group: monocyte count <363/mm3, according to the cut-off value calculated by median values). Table 2 shows the baseline characteristics of patients with non-HFpEF in the high and low peripheral monocyte groups. The prevalence of men, atrial fibrillation, diuretics, and New York Heart Association (NYHA) III or IV were significantly higher in the high peripheral monocyte group (P=0.005, P=0.02 and P=0.05, respectively; Table 2). These patients had lower eGFR (P=0.04; Table 2), and higher high-sensitivity C-reactive protein (hs-CRP; P=0.003; Table 2). Echocardiography parameters, such as LVEF, E/e′, SVI, TRPG, PAP, CO, and plasma BNP levels, did not differ between these 2 groups.
Baseline Characteristics of Non-HFpEF Patients With Peripheral Monocyte Count Measurements
| High monocyte group (monocytes >363; n=103) |
Low monocyte group (monocytes <363; n=97) |
P value | |
|---|---|---|---|
| Age (years) | 67.2±12.9 | 67.6±11.8 | 0.8 |
| Sex, male (%) | 84 (81.6) | 64 (66) | 0.005 |
| SBP (mmHg) | 122.1±25.7 | 121.1±18.6 | 0.7 |
| DBP (mmHg) | 72.2±18.2 | 70.7±13.2 | 0.5 |
| BMI (kg/m2) | 23.7±4.0 | 23.2±4.0 | 0.4 |
| Hypertension (%) | 71 (68.9) | 60 (61.9) | 0.2 |
| Diabetes (%) | 41 (39.8) | 44 (45.4) | 0.6 |
| Dyslipidemia (%) | 65 (63.1) | 59 (60.8) | 0.6 |
| Atrial fibrillation (%) | 36 (35) | 20 (20.6) | 0.02 |
| ACE-I or ARB (%) | 88 (85.4) | 82 (84.5) | 0.5 |
| CCB (%) | 34 (33) | 30 (30.9) | 0.6 |
| β-blocker (%) | 90 (87.4) | 87 (89.7) | 1.0 |
| Diuretics (%) | 75 (72.8) | 56 (57.7) | 0.01 |
| CAD (%) | 53 (51.5) | 53 (54.6) | 0.9 |
| NYHA III or IV (%) | 32 (31.1) | 19 (19.6) | 0.05 |
| eGFR (mL/min/1.73 m2) | 55.6±19.2 | 61.2±18.4 | 0.04 |
| Hemoglobin (g/dL) | 13.3±2.4 | 13.7±5.5 | 0.5 |
| Peripheral monocyte counts (n/mm3) | 494.8±144 | 272±67 | 0.003 |
| hs-CRP (ng/dL) | 0.57±1.3 | 0.16±0.4 | 0.003 |
| BNP (pg/mL) | 317±378 | 249±384 | 0.2 |
| LVEF (%) | 38±8.8 | 40.5±7.0 | 0.1 |
| LAD (mm) | 40.8±7.5 | 40.2±7.2 | 0.5 |
| E/e′ | 15.2±6.8 | 16.2±8.5 | 0.4 |
| SVI (mL/m2) | 32.0±11.0 | 33.5±9.0 | 0.3 |
| TR-PG (mmHg) | 25.9±12 | 25.4±11 | 0.7 |
| PAP (mmHg) | 32.5±13.4 | 31.3±11.4 | 0.5 |
| CO (L/min) | 3.60±1.04 | 3.61±1.05 | 0.9 |
Data are presented as the mean±SD, or median (%). Abbreviations as in Table 1.
Correlation Between Peripheral Monocyte Counts and Other Biomarkers in Non-HFpEF Patients
The correlation between peripheral monocyte counts and other biomarkers in patients with non-HFpEF was also investigated. In patients with non-HFpEF, peripheral monocyte counts correlated significantly and positively with hs-CRP (r=0.31; P<0.001); however, they did not correlate with Ln-BNP, a representative biomarker of cardiac overload (r=0.04; P=0.50; Supplementary Figure 1).
Cox Proportional Hazard Analysis of HF-Related Events in Non-HFpEF PatientsMultivariate Cox hazard analysis identified high peripheral monocyte counts as an independent and significant predictor of future HF-related events in patients with non-HFpEF (HR 2.24; 95% CI 1.01–4.99; P=0.04; Table 3). We included the following variables: NYHA III or IV and atrial fibrillation in forced 1; Ln-BNP and LVEF and diuretics in forced 2; and age, eGFR and hemoglobin in forced 3.
Cox Hazard Analysis of HF-Related Events in Non-HFpEF Patients
| Variable | Univariate regression | Multivariate regression | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Forced inclusion 1 | Forced inclusion 2 | Forced inclusion 3 | ||||||||||
| HR | 95% Cl | P value | HR | 95% Cl | P value | HR | 95% Cl | P value | HR | 95% Cl | P value | |
| NYHA III or IV (yes) |
3.02 | 1.46–6.27 | 0.003 | 2.26 | 1.05–4.84 | 0.03 | – | – | – | – | – | – |
| Age (years) | 1.04 | 1.005–1.081 | 0.02 | – | – | – | – | – | – | 1.01 | 0.97–1.05 | 0.4 |
| Male (yes) | 0.86 | 0.39–1.92 | 0.72 | – | – | – | – | – | – | – | – | – |
| Hypertension (yes) |
0.74 | 0.35–1.57 | 0.4 | – | – | – | – | – | – | – | – | – |
| Diabetes (yes) | 0.69 | 0.32–1.48 | 0.69 | – | – | – | – | – | – | – | – | – |
| Dyslipidemia (yes) |
0.65 | 0.31–1.36 | 0.25 | – | – | – | – | – | – | – | – | – |
| Atrial fibrillation (yes) |
2.82 | 1.36–5.86 | 0.005 | 2.11 | 0.98–4.51 | 0.054 | – | – | – | – | – | – |
| IHD (yes) | 1.05 | 0.5–2.19 | 0.8 | – | – | – | – | – | – | – | – | – |
| eGFR (mL/min/1.73 m2) |
0.96 | 0.94–0.98 | <0.001 | – | – | – | – | – | – | 0.98 | 0.96–1.00 | 0.06 |
| Hemoglobin (g/dL) |
0.67 | 0.57–0.78 | <0.001 | – | – | – | – | – | – | 0.74 | 0.63–0.87 | <0.001 |
| Ln-BNP | 2.71 | 1.90–3.87 | <0.001 | – | – | – | 2.52 | 1.67–3.79 | <0.001 | – | – | – |
| hs-CRP (ng/dL) | 1.31 | 1.12–1.53 | 0.001 | – | – | – | – | – | – | – | – | – |
| LVEF (%) | 0.95 | 0.92–0.99 | 0.03 | – | – | – | 1.02 | 0.97–1.07 | 0.2 | – | – | – |
| ACE-I or ARB (yes) |
0.53 | 0.21–1.31 | 0.17 | – | – | – | – | – | – | – | – | – |
| CCB (yes) | 0.66 | 0.28–1.54 | 0.33 | – | – | – | – | – | – | – | – | – |
| β-blocker (yes) | 1.02 | 0.31–3.38 | 0.96 | – | – | – | – | – | – | – | – | – |
| Diuretics (yes) | 4.61 | 1.39–15.2 | 0.01 | – | – | – | 2.11 | 0.6–7.3 | 0.2 | – | – | – |
| High peripheral monocyte count (yes) |
3.87 | 1.57–9.52 | 0.003 | 2.41 | 1.08–5.37 | 0.03 | 2.24 | 1.01–4.99 | 0.04 | 2.83 | 1.24–6.44 | 0.01 |
CI, confidence interval; HR, hazard ratio; IHD, ischemic heart disease; Ln-BNP, natural logarithmic transformed B-type natriuretic peptide. Other abbreviations as in Table 1.
ROC Analysis for Peripheral Monocyte Counts and Ln-BNP to Predict Occurrence of HF-Related Events in Non-HFpEF Patients
To assess the incremental prognostic ability of monocytes, ROC analysis was performed for the logistic regression model of Ln-BNP only, peripheral monocyte counts only, and peripheral monocyte counts plus Ln-BNP. The AUC curve for detection of HF-related events in patients with non-HFpEF were 0.85 (95% CI 0.79–0.92; P<0.001; Figure 2A), 0.67 (95% CI 0.56–0.78; P=0.003; Figure 2B), and 0.87 (95% CI 0.80–0.94; Figure 2C), respectively.
Receiver operating characteristic curves for peripheral monocyte count plus Ln-BNP level identifying heart failure (HF)-related events in non-HFpEF patients. AUC, area under the curve; HFpEF, heart failure with preserved left ventricular ejection fraction; Ln-BNP, natural logarithmic transformed B-type natriuretic peptide.
Harrell’s C-Statistic for Regression Model, Continuous NRI and IDI
The Harrell’s C-statistic value for Ln-BNP was 0.853 (95% CI 0.79–0.92); after adding peripheral monocyte counts as a factor, the value was 0.868 (95% CI 0.80–0.94; P=0.314). We reclassified the risk of HF-related events after adding peripheral monocyte counts to Ln-BNP; the continuous NRI was 73.6% (P<0.001), and the IDI was 4.5% (P=0.014; Supplementary Table 2).
DCA of Ln-BNP and Ln-BNP Plus Peripheral Monocyte CountsWe analyzed the DCA for Ln-BNP and Ln-BNP plus peripheral monocyte counts to predict 1,500 days of HF-related events in patients with non-HFpEF. Supplementary Figure 1 shows the DCA for Ln-BNP and Ln-BNP plus peripheral monocyte counts, and both model’s areas were useful between 0 and 40%. However, the combined model was superior to the Ln-BNP for a threshold probability of 40–60% (Supplementary Figure 2).
The primary findings of this study were as follows: (1) the Kaplan-Meier curve revealed a significantly higher probability of HF-related events in patients with high peripheral monocyte counts (≥363/mm3) only in non-HFpEF, and not in patients with HFpEF; (2) multivariate Cox proportional hazard analysis revealed that a high peripheral monocyte count was a significant and independent predictor of HF-related events in patients with non-HFpEF; (3) the NRI was significantly improved after adding peripheral monocyte counts to Ln-BNP in patients with non-HFpEF; (4) in the DCA for Ln-BNP and Ln-BNP plus peripheral monocyte counts, while both model’s areas were useful between 0 and 40%, the combined model was superior to the Ln-BNP for a threshold probability of 40–60% in patients with non-HFpEF.
The presence of chronic HF is associated with an increased risk of death, and patients with HF have a poor prognosis. Heart failure readmissions and outpatient worsening heart failure events are associated with worse clinical outcomes and remain a significant therapeutic challenge.17–19 However, the relationship between leukocyte subtype counts and the prognosis of patients with HF remains unclear. In this study, the occurrence of HF-related events was significantly higher in the high monocyte counts group than in the low monocyte counts group in patients with non-HFpEF, and high peripheral monocyte counts was a significant and independent predictor of HF-related events in patients with non-HFpEF patients. HF is a disease of long-term and persistent cardiac injury where the loss of myocardium is multifactorial. Common and easily recognized causes include CAD and hypertension, and increased levels of monocyte activation have been demonstrated in HF.20 Monocytes appear to be more than simply a marker of inflammation, they play a role in directly influencing the disease process. They are attracted to the failing myocardium and the expression of cell adhesion molecules, critical to monocyte recruitment and homing, increase with the severity of HF, and are associated with clinical outcomes.21,22 Monocytes are heavily involved in tissue damage and repair, and an imbalance of these processes may eventually lead to progressive adverse remodeling, interstitial fibrosis, and impaired contractility.20 This study demonstrates the utility of monocyte counts as a prognostic marker only in non-HFpEF, which is consistent with previous studies.
In this study, monocyte counts significantly correlated with HF-related events in patients with non-HFpEF and not in patients with HFpEF, indicating that the effects of inflammation on HF pathophysiology might vary depending on the category of HF. HFpEF is preceded by chronic comorbidities, such as hypertension, type 2 diabetes, obesity, and renal insufficiency; in contrast, non-HFpEF is often preceded by the acute or chronic loss of cardiomyocytes due to ischemia, a genetic mutation, myocarditis, or valvular disease.23,24 In non-HFpEF, systemic and cardiac inflammation are secondary causes of cardiomyocyte loss.25 Maekawa et al. reported that the maximum monocyte count after MI is 900/mm3 or more, and major cardiovascular events such as long-term cardiac death, hospitalization for HF, and reinfarction occur at a high rate.26 They also revealed that administration of granulocyte-macrophage colony-stimulating factor to a rat infarct model increased the expression of monocyte mobilization activity factor in the infarcted myocardium and the macrophage immersion along with an increase in peripheral monocyte counts. This worsens the remodeling of LV by suppressing the repairable fibrosis of the infarcted area.27
Conversely, HFpEF is characterized by low-grade chronic systemic inflammation and capillary dysfunction, with consequential low-grade cardiac inflammation. Our finding that monocyte counts were not related to HF-related events in patients with HFpEF indicates that chronic low-grade systemic inflammation in HFpEF might not be reflected in blood monocyte counts.28 Thus, we speculate that the different prognoses associated with the type of HF and peripheral monocyte counts in this study are related to different mechanisms of the onset and progression of inflammation and various involvement of monocytes in each category of HF.
HFpEF and non-HFpEF differ fundamentally in the role and impact of inflammation on disease progression. In non-HFpEF, inflammation is predominantly triggered by acute or chronic myocardial injury, such as ischemia, myocarditis, or valvular dysfunction. These injuries activate monocytes, which infiltrate the myocardium and contribute to adverse remodeling through processes such as fibrosis and extracellular matrix deposition. Elevated peripheral monocyte counts in non-HFpEF likely reflect this systemic inflammatory state, making monocyte count a reliable prognostic marker. However, this study focused on overall peripheral monocyte counts rather than specific monocyte subpopulations, which limits its ability to capture the nuanced roles of different monocyte subsets in HF progression. Studies such as the one by Elchinova et al. have highlighted the prognostic value of specific monocyte subsets (classical, intermediate, and non-classical monocytes) in HF patients.29 These subsets may play distinct roles in inflammation and myocardial remodeling, and their analysis could provide deeper insights into the pathophysiology of HF.
In contrast, HFpEF is characterized by chronic low-grade systemic inflammation, which is driven by systemic comorbidities such as hypertension, diabetes, and obesity. These conditions lead to endothelial dysfunction and microvascular inflammation, which in turn exacerbate myocardial stiffness and diastolic dysfunction. Unlike non-HFpEF, the inflammatory response in HFpEF is more localized and may not significantly increase peripheral monocyte counts. Instead, HFpEF involves complex interactions between endothelial cells, immune cells, and pro-inflammatory cytokines, such as interleukin (IL)-6 and tumor necrosis factor (TNF)-α, which promote capillary rarefaction and myocardial fibrosis. Future studies should incorporate advanced techniques, such as flow cytometry, to investigate the specific roles and contributions of monocyte subsets, such as classical, intermediate, and non-classical monocytes, in HF progression, with a particular focus on their distinct roles in HFpEF and non-HFpEF, building on the foundation laid by the present study.
Additionally, while Greene et al. demonstrated the prognostic utility of monocyte counts in HFrEF, their study did not explore the incremental value of combining monocyte counts with other biomarkers, such as BNP. Our findings address this gap by showing that the combination of monocyte counts and Ln-BNP significantly improves risk stratification in non-HFpEF patients. Furthermore, by incorporating a direct comparison between HFpEF and non-HFpEF, we highlight the limited predictive value of monocyte counts in HFpEF, potentially due to the distinct pathophysiological role of chronic low-grade inflammation in this phenotype.30 These differences emphasize the need for phenotype-specific approaches in heart failure research and clinical practice. The integration of advanced imaging techniques and molecular biomarkers could provide deeper insights into the localized inflammatory processes in HFpEF, as compared with systemic inflammation in non-HFpEF. These approaches may help clarify the distinct roles of inflammation and monocyte dynamics in different HF phenotypes. Understanding these differences could inform targeted therapeutic strategies that address the specific inflammatory pathways in HFpEF and non-HFpEF.
Findings on HF outcomes show that most attempts to target inflammation in the HF setting in phase III clinical trials resulted in neutral effects or worsening of clinical outcomes.31 In the Canakinumab Anti-Inflammatory Thrombosis Outcomes Study (CANTOS) trial, targeted anti-cytokine therapy with a monoclonal antibody against IL-1β resulted in the improvement of HF outcomes in patients with MI with or without established HF.32 Despite the past failures of inflammation-targeted therapies in heart failure, such as the HERMES trial, efforts to target this mechanism continue. These failures may be attributed to previous approaches relying on generalized immunomodulation without focusing on specific inflammatory pathways or immune cell subsets. Recent studies have shown promising results by targeting cytokine pathways or specific immune cell subsets. For instance, the CANTOS trial demonstrated that IL-1β inhibition could reduce cardiovascular events in patients with myocardial infarction. These findings underscore the importance of identifying precise inflammatory targets and understanding the distinct inflammatory mechanisms in HFpEF and non-HFpEF. Future research should focus on developing targeted therapies based on these insights to improve the outcomes of heart failure patients. The immune system acutely responds to both HF decompensation and its treatment.33 Thus, peripheral monocyte counts might fluctuate during the treatment process, as well as other inflammatory markers like CRP and IL-6.34 Although monocytes play a central role in the inflammation and pathophysiology of HF, many questions remain regarding their exact mechanism of action and the significance of different monocyte subpopulations. Using flow cytometry, different monocyte subsets can be defined by the relative expressions of CD14 and CD16.35 These subsets may be viewed as separate entities because they appear to have distinct functionality.20 One study explored the distribution of monocyte subsets in patients with stable HF compared with healthy controls and showed that the CD14_CD16_ subset was increased in patients with HF, while the CD14 low CD16_ subpopulation was depleted.36 Given the distinct function of each monocyte subpopulation, each subset may participate in the inflammatory pathophysiology of HF to varying degrees. To date, efforts to therapeutically target monocyte-derived inflammatory cytokine systems in HF have been unsuccessful,37,38 differing from neurohormonal antagonists for heart failure patients.39 However, it may be more effective to replace the generalized immunomodulation approach with efforts to target specific monocyte subsets. To elucidate the difference in monocyte involvement between HFpEF and non-HFpEF, large-scale prospective studies targeting inflammation will be required not only in patients with non-HFpEF, such as the CANTOS trial, but also in patients with HFpEF.
Study LimitationsThere are some limitations in this study. First, this was a 1-center study with a relatively small patient population. However, even in this small population, the monocyte count was closely associated with the prognosis of patients with non-HFpEF. Further prospective multicenter studies involving a larger patient population will be required to determine the importance of leukocyte subtype counts in patients with HF. Second, this study was observational and did not include an intervention such as anti-inflammatory drugs. The benefits of anti-inflammatory therapy for cardiovascular diseases remain unclear. Therefore, prospective interventional studies in patients with non-HFpEF in a large-scale population are necessary. Third, in this study, we measured hs-CRP; however, we did not measure pro-inflammatory mediators, such as monocyte chemotactic protein-1 and TNF-α. Hence, we could not examine the relationship between monocyte counts and other inflammatory biomarkers and their involvement in LV systolic dysfunction. Fourth, HF-related events were defined solely as hospitalizations for HF decompensation, and sudden death or cardiovascular death was not included in the analysis. Fifth, studies such as the one by Elchinova et al., have highlighted the prognostic value of specific monocyte subsets (classical, intermediate, and non-classical monocytes) in HF patients.27 We focused on clinical significance of monocyte count as a biomarker for the prognosis of HF patients, because monocyte counts in peripheral blood are easily measurable and widely available biomarker in clinical practice. Although this approach has its limitations in capturing the complexities of immune cell dynamics and molecular mechanisms of inflammation in HF patients, we believe that the present study provides clinical significance of monocyte counts in HF prognosis, especially in distinguishing between HFpEF and non-HFpEF. We acknowledge that detailed analyses using advanced techniques, such as flow cytometry, would allow for a more nuanced evaluation of the specific roles of monocyte subsets in HF pathophysiology, and future studies should aim to build on our findings by incorporating detailed analyses of monocyte subsets to further clarify their prognostic and mechanistic roles in HF. This approach may have introduced competing risks, such as death, which could potentially influence the interpretation of our findings. While we used Cox proportional hazard models for the survival analysis, competing risks were not explicitly accounted for. The Fine and Gray subdistribution hazard model, which is designed to address competing risks, could provide additional insights and more accurate interpretations. Future studies incorporating such models are warranted to further validate our results.
The monocyte count predicts HF-related events only in patients with non-HFpEF, suggesting differences in the involvement of inflammation between HFpEF and non-HFpEF.
None.
None.
K.T. is a member of Circulation Reports’ Editorial Team.
All procedures were conducted in accordance with the Declaration of Helsinki and its amendments. The study protocol was approved by the institutional review board of Kumamoto University (approval no. Senshin 2225). This study is registered with the University Hospital Medical Information Network Clinical Trials Registry (UMIN000038771).
Public access information: opt-out materials are available at http://www.kumadai-junnai.com/home/wp-content/uploads/houkatsu.pdf.
Please find supplementary file(s);
https://doi.org/10.1253/circrep.CR-25-0102
