Incremental Prognostic Value of Platelet Count in Patients With Acute Heart Failure　― A Retrospective Observational Study ―

Abstract

Background: Acute heart failure (AHF) triggers platelet aggregation and platelet markers are associated with the severity of AHF. The present study aimed to investigate the prognostic value of platelet count (PLT) in patients with AHF.

Methods and Results: This single-center retrospective observational study analyzed 425 consecutive patients with AHF. The patients were divided into groups based on tertiles of PLT: low (PLT1 <170,000/μL), intermediate (170,000/μL≤PLT<230,000/μL), and high (PLT3 ≥230,000/μL). The endpoint was all-cause death with a composite endpoint of all-cause death and HF rehospitalization. Survival analysis was performed, and Cox proportional hazard models adjusted by an established risk score (Get With The Guidelines score) were generated. The PLT1 group had the worst survival for all-cause death (log-rank, P=0.003) and the composite endpoint (P=0.009). A significant trend of increasing survival was observed for all-cause death (log-rank trend, P<0.001) and the composite endpoint (P=0.002) in the following order: PLT1, PLT2, and PLT3. Adjusted Cox proportional hazard models demonstrated that low PLT was a risk factor of all-cause death and the composite endpoint.

Conclusions: Low PLT was associated with risk for all-cause death and HF rehospitalization in patients with AHF.

Acute heart failure (AHF) is characterized by acute deterioration of congestive symptoms and pump failure caused by structural and functional cardiac dysfunction.^1,2 The severity of symptoms varies from mild to lethal. Because the pathophysiological background and development of AHF are complicated and heterogeneous,³ accurate assessment of severity remains a challenge. Several risk biomarkers have been reported: B-type natriuretic peptide (BNP) as a marker of left ventricular stretch and congestive state,⁴ C-reactive protein (CRP) and the neutrophil to lymphocyte (N/L) ratio as markers of inflammation,^5,6 and blood urea nitrogen (BUN) as a marker of renal function and renin-angiotensin-aldosterone (RAA) system activation.⁷

Editorial p 511

AHF has been reported to trigger platelet activation and aggregate formation.⁸ Several studies have demonstrated that platelet activation and pro-thrombotic state are related to the severity of AHF and the potential risk of HF deterioration in patients with chronic HF. Ueland et al demonstrated that a platelet marker was associated with HF congestive symptoms,⁹ and the mean platelet volume (MPV), which represents platelet activation, was related to the risk of HF hospitalization in patients with chronic HF.¹⁰ Meanwhile, age-related reduction in the platelet count (PLT) has been reported in a healthy population,¹¹ but bone marrow dysfunction, independent of aging, has been reported in symptomatic chronic HF.¹² Furthermore, in patients newly diagnosed with HF with reduced ejection fraction (EF <40%), thrombocytopenia (<100,000/μL) was a risk factor of 1-year mortality.¹³ Collectively, the prognostic value of PLT in AHF remains unclear.

In the present study we aimed to investigate the prognostic value of PLT on admission in patients with AHF, the relationship between PLT and other biomarkers (MPV, BNP, CRP, N/L ratio, and BUN) and age in AHF, and the change in PLT between 30 days before hospitalization and admission.

Methods

Participants

This single-center retrospective study at a Japanese community hospital enrolled 469 consecutive patients admitted to the cardiology ward because of AHF. At least 2 cardiologists certified by the Japanese Circulation Society diagnosed AHF in the clinic or emergency room between June 2014 and July 2016.¹⁴ The diagnosis of AHF was based on the Framingham criteria.¹⁵ All patients had HF exacerbation with New York Heart Association class III or IV and at least one of the following signs of congestion: pulmonary edema, pitting edema in the lower extremities, distended jugular vein, and/or pleural effusion. None of the patients required a cardiac support device, such as intra-aortic balloon pump or left ventricular assist device for cardiogenic shock, nor had they concurrent acute myocardial infarction on admission. No cases of liver cirrhosis were detected in the study population. We excluded 44 patients because their PLT was not measured, so a total of 425 patients were eligible for analysis (Figure 1).

Figure 1.

Flowchart of the enrollment. PLT, platelet count.

Prior to the allocation, we planned 3 group comparisons to confirm the trend of survival. The 1-year mortality rate was estimated to be 10% in the lowest mortality group and 25% in the highest mortality group.¹⁶ With a power of 0.8 and α error of 0.05, each group required ≥78 patients. Considering feasibility, the patients were divided into 3 groups based on tertiles of PLT: low (PLT1 <170,000/μL), intermediate (170,000/μL≤PLT<230,000/μL), and high (PLT3 ≥230,000/μL).

The study was conducted in accordance with the Declaration of Helsinki and approved by the local ethical committee at Tomishiro Central Hospital. Informed consent was waived because of the observational nature of the study. For confidentiality, all data were de-identified and analyzed anonymously.

Data Collection

We conducted an in-depth review of the medical chart to collect both demographic and laboratory data.

All blood samples were obtained from the brachial veins and stored in a test tubes (Venoject VP-DK052K05, Terumo, Japan). After sampling, blood cells were immediately counted using XE2100 (Sysmex Corp., Kobe, Japan).

Based on the PLT on admission and at 30 days before hospital admission (baseline PLT) if available, we calculated the percent change in PLT: [(PLT on admission)−(baseline PLT)]/(baseline PLT)×100.

Statistical Analysis

All missing data were checked. The normality of the continuous variables was assessed using histograms. Continuous variables with normal and skewed distribution were expressed as mean±SD and median [25%, 75%], respectively. Categorical variables were expressed as number (%). BNP, CRP, and BUN had highly right-skewed distributions and were log-transformed (logBNP, logCRP, and logBUN) to achieve normal distribution. The demographic characteristics and laboratory data on admission were summarized, and data on PLT1, PLT2, and PLT3 were compared. One-way analysis of variance and Kruskal-Wallis test were used for the comparison of normally distributed and non-normally distributed continuous variables, and Fisher’s exact test was used for categorical variables. Posthoc analysis for the comparisons among PLT3 and PLT1 and PLT3 and PLT2 was performed using Holm’s test.

Survival Analysis Survival analysis was performed for all-cause death in 425 patients (Figure 1). Time zero was the date of hospital admission, and observation was censored at all-cause death as the event or last hospital visit without an event. To compare survival among the 3 groups, Kaplan-Meier curves were generated, and log-rank test and log-rank trend test were performed to test the trend in the following order: PLT1, PLT2, and PLT3. Posthoc analysis was performed for comparison among the 3 groups using Holm’s test.

Cox Proportional Hazard Model We generated univariate Cox hazard regression models for all-cause death and Cox cause-specific proportional hazard models for the composite endpoint to compute the hazard ratios with 95% confidence intervals. Each of the following 8 factors was included in a separate model: the Get With The Guidelines (GWTG) score,^17,18 an established risk score; PLT (PLT1 to PLT3 and PLT2 to PLT3); biomarkers including MPV,¹⁰ logBNP,⁴ logCRP,⁵ N/L ratio,⁶ and logBUN.¹⁹ Each of 6 factors other than the GWTG score and BUN was entered into Cox hazard models adjusted by GWTG score. BUN was taken into account for GWTG score calculation. For these models, cases with missing data for the variables of interest were excluded (list-wise deletion).

Correlation Between PLT and Biomarkers Pearson’s or Spearman’s correlation coefficient was used to analyze the relationship between PLT and the following biomarkers: MPV, logBNP, logCRP, N/L ratio, and logBUN.

Correlation Between PLT and Age Pearson’s correlation coefficient was used to evaluate the relationship between PLT and age.

Subgroup Analysis for Change in PLT The patients were subdivided accordingly to baseline PLT. A total of 217 patients were eligible for subgroup analysis. We compared the baseline PLT and PLT on admission using Wilcoxon’s signed-rank test in the overall subgroup population and in each group.

All analyses used R 3.4.3 (R Foundation for Statistical Computing, Vienna, Austria) and EZR 1.3.6 (Saitama Medical Center, Jichi Medical University, Saitama, Japan).²⁰ All reported P values are two-tailed, and P<0.05 was considered statistically significant.

Results

Participants

The mean age of the participants was 79±12 years, with 202 (48%) male participants (Table 1 and Supplementary Table). The mean GWTG score was 38±7. The PLT1 group consisted of elderly patients with higher GWTG than the PLT3 group. Furthermore, the PLT1 group had a higher BUN, aspartate aminotransferase, creatinine, total bilirubin, and MPV and lower estimated glomerular filtration rate than the PLT3 group.

Table 1. Patients’ Baseline Characteristics

	Overall (n=425)	PLT3 (n=141)	PLT2 (n=142)	PLT1 (n=142)	P value
Demographic data
Age, years	79±12	77±13	79±13	81±9.7	0.009
Male	202/425 (48)	59/141 (42)	64/142 (45)	79/142 (56)	0.053
Body weight, kg	58±145	58±14	59±16	57±15	0.36
Body mass index, kg/m2	22.7±4.6	22.9±4.1	23.1±5.0	22.2±4.5	0.21
Body surface area, m2	1.51±0.22	1.50±0.20	1.53±0.23	1.50±0.22	0.43
Height, cm	154±10	154±9	155±10	155±10.02	0.50
GWTG score	38±7	37±7	38±8	40±7	0.001
Hospital stay, days	13 [9, 21]	14 [10, 19]	13 [8, 19]	14 [9, 25]	0.33
LVEF, %	46±17	45±17	46±17	45±16	0.98
Past medical history, n (%)
OMI	75/425 (18)	31/141 (22)	25/142 (18)	19/142 (13)	0.17
AF	122/425 (29)	42/141 (30)	38/142 (27)	42/142 (30)	0.82
Hemodynamic data
Systolic BP, mmHg	131±26	135±26	131±28	128±23	0.086
Heart rate, beats/min	83±20	85±20	84±21	80±19	0.065
Laboratory data
BUN, mg/dL	24.0 [17.0, 35.3]	22.0 [16.0, 30.0]	24.0 [18.0, 33.0]	31.0 [18.0, 43.0]	<0.001
Creatinine, mg/dL	1.15 [0.82, 1.54]	1.06 [0.76, 1.46]	1.15 [0.83, 1.44]	1.31 [0.90, 2.01]	0.001
eGFR, mL/min/1.73/m2	44.9±22.8	49.3±24.4	44.7±20.4	40.6±22.6	0.006
Albumin, g/dL	3.5±0.5	3.5±0.6	3.6±0.5	3.5±0.6	0.55
AST, mg/dL	23 [17, 32]	21 [17, 27]	24 [18, 32]	25 [17, 41]	0.011
ALT, mg/dL	17 [12, 29]	17 [11, 27]	18 [12, 28]	18 [12, 33]	0.50
AST/ALT	1.36 [0.97, 1.80]	1.33 [0.93, 1.67]	1.27 [0.92, 1.83]	1.45 [1.10, 2.00]	0.13
Total bilirubin, mg/dL	0.6 [0.4, 1.0]	0.50 [0.40, 0.70]	0.60 [0.40, 1.00]	0.80 [0.50, 1.10]	<0.001
PLT on admission, 104/mcL	20.6±8.0	29.2±6.4	19.9±1.6	12.6±3.4	<0.001
Baseline PLT, 104/mcL	19.6±7	25.5±7.4	19.6±4.2	14.7±4.4	<0.001
Change in PLT (%)	4.9±31.9	21.5±32.2	4.4±23	–8.2±32	<0.001
Mean platelet volume, fL	10.7±1.1	10.26 (0.84)	10.56 (1.00)	11.10 (1.16)	<0.001
Hemoglobin, g/dL	11.9±2.5	11.6±2.5	12.2±2.4	12.0±2.5	0.20
N/L	3.29 [2.30, 5.05]	3.28 [2.37, 4.55]	3.60 [2.33, 5.51]	2.99 [2.12, 4.95]	0.68
BNP, pg/mL	625 [381, 1,031]	505 [312, 863]	709 [386, 973]	645 [405, 1,246]	0.22
CRP, mg/dL	1.64 [0.53, 4.44]	1.56 [0.58, 4.61]	1.68 [0.37, 3.73]	1.60 [0.60, 5.37]	0.38
Na, mEq/L	139±5	138±6	139±5	138±5	0.48
Echocardiographic parameter
IVC diameter, mm	17.2±4.9	16.4±5.1	17.5±4.7		0.065
Medications, n (%)
Aspirin	4/425 (11)	10/141 (7.1)	18/142 (13)	16/142 (11)	0.27
Clopidogrel	13/425 (3.1)	3/141 (2.1)	7/141 (4.9)	3/142 (2.1)	0.32
DOAC	16/425 (3.8)	7/141 (5)	4/142 (2.8)	5/142 (3.5)	0.60
Warfarin	28/425 (6.6)	8/141 (5.7)	8/142 (5.6)	12/142 (8.5)	0.57
ACEI and/or ARB	124/425 (29)	46/141 (33)	43/142 (30)	35/142 (25)	0.31
β-blocker	153/425 (36)	51/141 (36)	50/142 (35)	52/142 (37)	0.98

One-way analysis of variance and Kruskal-Wallis test were used for continuous variables with normal and skewed distribution for comparison among the 3 groups. Fisher’s exact test was used for categorical variables for comparison among the 3 groups. Participants were subdivided according to PLT (low as PLT1, intermediate as PLT2, and high as PLT3). Results of posthoc analysis are shown in Supplementary Table. ACEI, angiotensin-converting enzyme inhibitor; AF, atrial fibrillation; ALT, alanine aminotransferase; ARB, angiotensin receptor blocker; AST, aspartate aminotransferase; BNP, B-type natriuretic peptide; BP, blood pressure; BUN, blood urea nitrogen; CRP, C-reactive protein; DOAC, direct oral anticoagulants; eGFR, estimated glomerular filtration rate; GWTG, Get With The Guidelines score; IVC, inferior vena cava; LVEF, left ventricular ejection fraction; N/L, ratio of neutrophils to lymphocytes; OMI, old myocardial infarction; PLT, platelet count (baseline PLT defined as PLT 30 days prior to hospital admission).

Survival Analysis

During the follow-up (201 [92, 396] days), 100 (24%) patients died, and 86 (20%) patients were readmitted to hospital because of HF exacerbation.

As shown in Figure 2A and Supplementary Figure A there was a significant difference in the incidence of death among the 3 groups according to the Kaplan-Meier curves for the comparison of all-cause death (P=0.014 for the comparison of the 3 groups; PLT1 42/142 (30%) vs. PLT3 22/141 (16%), P=0.02) and survival (log-rank, P=0.003). Both the PLT1 and PLT2 groups had poor survival compared with the PLT3 group (PLT 1 vs. PLT3, P=0.003; PLT2 vs. PLT3, P=0.033). A significant trend of increasing survival for all-cause death was observed in the following order: PLT1, PLT2, and PLT3 (log-rank trend, P<0.001).

Figure 2.

Kaplan-Meier curves for (A) all-cause death and (B) composite endpoint of all-cause death and heart failure rehospitalization. Tertiles of platelet count: low (PLT1 <170,000/μL), intermediate (170,000/μL≤PLT<230,000/μL), and high (PLT3 ≥230,000/μL). Post-hoc analysis is shown in Supplementary Figure.

Figure 2B and Supplementary Figure B is the Kaplan-Meier curves for the comparison of composite endpoint among the 3 groups and no statistical difference was observed in the incidence of composite endpoint (PLT1 52/141 (37%), PLT2 63/142 (44%), PLT3 71/142 (50%); P=0.087 for the comparison of the 3 groups). However, a significant difference among the groups was observed in terms of survival (log-rank, P=0.009). The PLT1 group had poor survival compared with the PLT3 group (P=0.007). A significant trend of increasing survival was observed in the following order: PLT1, PLT2, and PLT3 (log-rank trend, P=0.002).

Cox Proportional Hazard Model for All-Cause Death

Univariate Cox proportional hazard models demonstrated that GWTG score, PLT1 and PLT2, increased logBNP, increased logCRP, elevated N/L ratio, and increased logBUN were significant risk factors for all-cause death in AHF (Table 2). Cox proportional hazard models adjusted by GWTG score also demonstrated that PLT1, increased logBNP, increased logCRP, and elevated N/L ratio were significant risk factors for all-cause death in AHF (Table 2).

Table 2. Cox Proportional Hazard Model for All-Cause Death

Factor	Univariate				Adjusted by GWTG
	Event/n	HR	95% CI	P value	Event/n	HR	95% CI	P value
GWTG score, 1 point increase each	98/419	1.1	1.07–1.13	<0.001	(Adjustment factor)
PLT2 to PLT2	100/424	1.83	1.08–3.12	0.026	98/419	1.71	0.99–2.94	0.054
PLT1 to PLT3	100/424	2.43	1.45–4.08	<0.001	98/419	2.02	1.19–3.42	0.01
Mean platelet volume, 1 fL increase each	23/147	1	0.69–1.45	0.99	23/145	0.99	0.71–1.37	0.93
LogBNP, 1 log (pg/mL) increase each	39/149	2.26	1.45–3.5	<0.001	37/145	1.81	1.14–2.88	0.012
LogCRP, 1 log (mg/dL) increase each	71/239	1.31	1.11–1.54	0.002	69/234	1.32	1.11–1.58	0.002
N/L, 1 increase each	37/147	1.18	1.08–1.28	<0.001	36/143	1.14	1.05–1.25	0.003
LogBUN, 1 log (mg/dL) increase each	98/419	2.81	1.84–4.29	<0.001	(Included in GWTG)

PLT on hospital admission was tertiled into control as PLT3, moderate depletion as PLT2, and severe depletion as PLT1. Abbreviations as in Table 1.

Cox Proportional Hazard Model for Composite Endpoint

Univariate Cox proportional hazard models demonstrated that GWTG score, PLT1, increased logBNP, and increased logBUN were significant risk factors for the composite endpoint in AHF (Table 3). Cox proportional hazard models adjusted by GWTG score also demonstrated that PLT1 and increased logBNP were significant risk factors for the composite endpoint in AHF (Table 3).

Table 3. Cox Proportional Model for Composite Endpoint of All-Cause Death and Rehospitalization for Heart Failure Exacerbation

Factor	Univariate				Adjusted by GWTG
	Event/n	HR	95% CI	P value	Event/n	HR	95% CI	P value
GWTG score, 1 point increase each	184/420	1.07	1.05–1.09	<0.001	(Adjustment factor)
PLT2 to PLT3	186/425	1.36	0.94–1.97	0.1	184/420	1.32	0.91–1.91	0.15
PLT1 to PLT3	186/425	1.78	1.24–2.55	0.002	184/420	1.53	1.06–2.21	0.023
Mean platelet volume, 1 fL increase each	41/148	0.76	0.56–1.03	0.082	41/146	0.75	0.57–1.01	0.054
LogBNP, 1 log (pg/mL) increase each	74/150	1.83	1.33–2.52	<0.001	72/146	1.57	1.13–2.18	0.007
LogCRP, 1 log (mg/dL) increase each	118/239	0.99	0.88–1.11	0.83	116/234	0.98	0.87–1.11	0.72
N/L, 1 increase each	69/147	1.05	0.97–1.14	0.22	68/143	1.04	0.96–1.12	0.31
LogBUN, 1 log (mg/dL) increase each	184/420	2.07	1.54–2.79	<0.001	(Included in GWTG)

The participants were tertiled into PLT1 as severe platelet depletion, PLT2 as moderate depletion, and PLT3 as control. Abbreviations as in Table 1.

Correlations Between PLT, Biomarkers and Age

A significant negative correlation was found between PLT and MPV (Pearson’s r=−0.35, P<0.001) and between PLT and logBUN (Pearson’s r=−0.191, P<0.001) (Table 4).

Table 4. Correlations Between PLT and Prognostic Markers in Acute Heart Failure

Factor	n	Correlation	P value
PLT vs. mean platelet volume	148	r=−0.35	<0.001
PLT vs. logBNP	150	r=−0.152	0.063
PLT vs. logCRP	239	r=−0.04	0.54
PLT vs. N/L	147	ρ=−0.001	0.99
PLT vs. logBUN	420	r=−0.191	<0.001
PLT vs. IVC diameter	362	ρ=−0.09	0.088

logBNP, log-transformed BNP; logBUN, log-transformed BUN; logCRP, log-transformed CRP; r, Pearson’s correlation coefficient; ρ, Spearman’s correlation coefficient. Other abbreviations as in Table 1.

There was a significant negative correlation between age and PLT (r=−0.12, P=0.02; Figure 3).

Figure 3.

Relationship between age and platelet count in patients with acute heart failure. r, Pearson’s correlation coefficient.

Change in PLT Prior to Hospital Admission

In the overall population, the PLT1 and PLT2 groups had lower baseline PLT than the PLT3 group (Table 1).

In the subgroup analysis, there was a significant decrease from baseline PLT to PLT at admission in the PLT1 group (baseline PLT, 14.7±4.4/μL and admission PLT, 12.6±3.4/μL; P=0.002; Figure 4). In contrast, the PLT did not change before hospital admission in the PLT2 group (baseline PLT, 19.6±4.2/μL and admission PLT, 19.6±1.7/μL; P=0.77), and the PLT increased significantly prior to hospital admission in the PLT3 group (baseline PLT, 25.5±7.4/μL and admission PLT, 29.6±7/μL; P<0.001).

Figure 4.

Change in platelet count (PLT) prior to hospital admission for acute heart failure. PLT sampled 30 days prior to hospital admission and on admission for acute heart failure. Wilcoxon’s signed-rank test was used to compare PLT prior to hospital admission and on admission. Error bar shows mean±standard deviation (SD). Tertiles of PLT: low (PLT1 <170,000/μL), intermediate (170,000/μL≤PLT<230,000/μL), and high (PLT3 ≥230,000/μL).

Discussion

To the best of our knowledge, this study is the first to show the incremental prognostic value of PLT in AHF. We have 5 major findings. First, low PLT was a risk factor for all-cause death and the composite endpoint of all-cause death and rehospitalization for HF. The PLT decreased in the PLT1 group. In contrast, the PLT increased in the PLT3 group. Second, a significant negative correlation was found between PLT and MPV. Third, a significant correlation was found between PLT and logBUN. Fourth, there was a significant negative correlation between PLT and age in AHF (Figure 4). Fifth, the PLT1 and PLT3 groups had different changes in PLT prior to hospital admission.

Prognostic Value of PLT

The current study found that low PLT is a risk factor for all-cause death and HF rehospitalization. Mojadidi et al demonstrated that thrombocytopenia (<100,000/μL) was associated with 1-year death in patients who were first diagnosed as having HF with reduced EF (<40%).¹³ Recent studies revealed that AHF was evoked not solely by cardiac problems but also by systemic physiological changes such as a pro-oxidant and inflammatory state.¹ The PLT may reflect those systemic changes and indicate the severity of AHF. Our study included patients with AHF with preserved and reduced EF, and showed that thrombocytopenia was associated with worse prognosis in such patients.

Previous studies reported that increased MPV was associated with worse prognosis in patients with HF, whereas PLT showed a borderline difference between patients with HF who died or survived within 30 days of admission.^21,22 Those studies included patients with stable HF who did not require hospitalization. We found a different change in PLT before hospitalization between the 1st and 3rd tertiles of PLT on admission. PLT may change according to the severity of HF exacerbation and had prognostic value in our study population.

The PLT1 and PLT2 groups had higher total bilirubin, but there was no significant difference in BNP between the groups (Table 1). However, this may reflect the preexisting long-standing congestive state resulting in congestive hepatopathy, not the short-term congestive state at admission, because BNP is thought to reflect only the short-term congestive state at admission, considering its half-life.²³ In contrast, chronic congestive hepatopathy decreases platelet production and increases platelet destruction.^24,25 Therefore, low PLT on hospital admission for HF might reflect long-standing congestion and coexisting congestive hepatopathy. The IVC diameter showed a trend of a negative correlation with PLT (Table 4). The finding of a relationship between BNP and PLT and with the IVC diameter indicated that the congestive state may be associated with a decrease in PLT to some extent. However, this trend was very weak. Multiple factors including congestion seem to contribute to the PLT change.

Correlation Between PLT and Other Biomarkers

A significant negative correlation was found between PLT and MPV. Platelets are held within cytoplasmic fragments of megakaryocytes and disk-shaped anucleate particles. The average life span of platelets is 8–10 days,^26,27 and their size does not change throughout their life span.^27,28 Moreover, megakaryocytopoiesis determines the size of platelets,^27,28 and an inverse relationship between PLT and MPV is generally observed with increased production of megakaryocytes.²⁹ Some physiological states such as inflammation, RAA activation, and platelet recruitment induce megakaryocytopoiesis.^27,30 These conditions coexist with AHF and may be accompanied by increased MPV and low PLT.

A significant negative correlation was found between PLT and logBUN, for which there are 2 possible explanations: RAA activation and renal function. BUN has been reported as a surrogate marker of RAA activation.⁷ Schäfer et al reported that angiotensin-converting enzyme inhibitors suppressed platelet activation in a rat HF model.³¹ In the present study, RAA activation might be the reason for the increase in BUN; in addition, RAA activation might cause platelet production, recruitment, aggregation, and consumption. In renal failure, BUN is increased,³² whereas the clearance of cytokines is decreased.³³ Worsening renal failure frequently occurs in patients with AHF.³⁴ In this study, the PLT1 group had higher serum creatinine and lower estimated glomerular filtration rate than the PLT3 group. Furthermore, the PLT1 group was presumed to have worse renal function than the PLT3 group. Impaired renal function might prolong pro-inflammatory cytokine clearance and induce platelet recruitment and consumption.

Relationship Between PLT and Age in AHF

An age-related reduction in the PLT was observed in our study population. The PLT1 group had older patients compared with the PLT2 and PLT3 groups (Table 1). The GWTG score is a summarized risk score that includes age; thus, the Cox proportional hazard model adjusted for GWTG score demonstrated PLT reduction was still a risk factor after adjustment by age. When focusing on the relationship between PLT and age (Figure 3), the negative correlation was weak (r=−0.12, P=0.02). PLT reduction in AHF seemed to be modulated more closely by other factors caused by HF such as a congestive state or platelet activation than age.

Change in PLT Prior to Hospital Admission

Subgroup analysis indicated that platelet depletion was not observed in all AHF patients. The PLT1 and PLT3 groups had different changes in PLT prior to hospital admission. This subgroup analysis had a considerable number of excluded patients because no data for the PLT 30 days prior to hospital admission were available (Figure 1). AHF comprises one-third de novo AHF and two-thirds exacerbation of chronic HF.¹ The subgroup population might not include de novo AHF but rather, exacerbation of chronic HF. In the patients with exacerbation of chronic HF, both the trend of decreasing PLT and low PLT on admission indicated worse prognosis.

PLT increased in the PLT3 group before hospital admission. Therefore, not only mechanisms for the decrease in PLT but also the increase in PLT seem to exist. The mechanism for the increase in PLT might be an inflammatory response manifesting as secondary thrombocytosis.³⁵ Moreover, the mechanisms for the decrease in PLT were presumed to be as follows: congestive state,²⁵ bone marrow dysfunction caused by congestion,¹² and sympathetic nerve activation.³⁶ The mechanisms for the decrease in PLT seemed to be closely related to HF exacerbation. Thus, severe AHF showed a decreased PLT.

Considering the finding of a significant different trend in PLT between the PLT1 and PLT3 groups prior to hospital admission, a change in PLT might help cardiologists better understand the HF condition in the management of chronic HF.

Clinical Significance

A doctor other than a cardiologist, such as an emergency doctor, might encounter patients with AHF and treat them in the primary care setting because AHF is quite common.³⁷ Blood cell count must be evaluated in all patients with AHF and the PLT is routinely measured in the clinical setting. In fact, there were fewer missing data for PLT than for the other biomarkers such as MPV, BNP, CRP, and N/L ratio in our study population. In the light of this observation, the use of the PLT is reasonable for risk stratification in AHF from the viewpoint of data availability.

Study Limitations

First, a racial difference in PLT has been reported,³⁰ and the present study only included Japanese patients. Therefore, generalization of findings with regard to other populations may be limited. Second, vascular resistance, cardiac structure, left ventricular aneurysm, and severe left ventricular wall motion abnormality, which might affect the PLT, were not examined.⁸ Third, we could not obtain information on coexisting liver disease because none of the patients in our study population had liver cirrhosis. Lastly, liver abnormality was not evaluated using ultrasonography or noninvasive indocyanine green clearance test. Thus, we could not provide information on liver function other than transaminase level.

Conclusions

In summary, platelet depletion may occur together with underlying pathophysiological changes including congestion and RAA activation in AHF. Low PLT was associated with poor prognosis in patients with AHF. PLT is widely measured in clinical settings and can be readily available for use as a risk marker in AHF.

Acknowledgments

We thank the following: Yoji Takami, Shimon Toma, and Chio Iseki for assisting in data collection and management; the medical technicians at Tomishiro Central Hospital for performing blood sampling and reporting laboratory data results; Atsushi Kakazu, Kazuaki Okuyama, Toshiya Chinen, Masanori Kakazu, Masahiro Tamashiro, Hideaki Sonoi, Hideki Takaesu, Masaki Tabuchi, and Akihiko Yamauchi for caring for the patients; Yumi Ikehara and Shoko Nakaima for assisting with writing this manuscript.

Funding

This study had no financial support.

Supplementary Files

Please find supplementary file(s);

http://dx.doi.org/10.1253/circj.CJ-18-0961

References

• 1.   Arrigo M, Parissis JT, Akiyama E, Mebazaa A. Understanding acute heart failure: Pathophysiology and diagnosis. Eur Heart J Suppl 2016; 18: G11–G18.
• 2.   Colombo PC, Doran AC, Onat D, Wong KY, Ahmad M, Sabbah HN, et al. Venous congestion, endothelial and neurohormonal activation in acute decompensated heart failure: Cause or effect? Curr Heart Fail Rep 2015; 12: 215–222.
• 3.   Joseph SM, Cedars AM, Ewald GA, Geltman EM, Mann DL. Acute decompensated heart failure: Contemporary medical management. Tex Heart Inst J 2009; 36: 510–520.
• 4.   Fonarow GC, Peacock WF, Phillips CO, Givertz MM, Lopatin M; ADHERE Scientific Advisory Committee and Investigators. Admission B-type natriuretic peptide levels and in-hospital mortality in acute decompensated heart failure. J Am Coll Cardiol 2007; 49: 1943–1950.
• 5.   Anand IS, Latini R, Florea VG, Kuskowski MA, Rector T, Masson S, et al. C-reactive protein in heart failure: Prognostic value and the effect of valsartan. Circulation 2005; 112: 1428–1434.
• 6.   Turfan M, Erdoğan E, Tasal A, Vatankulu MA, Jafarov P, Sönmez O, et al. Neutrophil-to-lymphocyte ratio and in-hospital mortality in patients with acute heart failure. Clinics 2014; 69: 190–193.
• 7.   Kazory A. Emergence of blood urea nitrogen as a biomarker of neurohormonal activation in heart failure. Am J Cardiol 2010; 106: 694–700.
• 8.   Chung I, Lip GYH. Platelets and heart failure. Eur Heart J 2006; 27: 2623–2631.
• 9.   Ueland T, Aukrust P, Yndestad A, Otterdal K, Frøland SS, Dickstein K, et al. Soluble CD40 ligand in acute and chronic heart failure. Eur Heart J 2005; 26: 1101–1107.
• 10.   Kaya H, Kutay Yıldırımlı M, Kurt R, Beton O, Birhan Yilmaz M. Mean platelet volume as a predictor of heart failure-related hospitalizations in stable heart failure outpatients with sinus rhythm. Acta Cardiol Sin 2017; 33: 292–300.
• 11.   Balduini CL, Noris P. Platelet count and aging. Haematologica 2014; 99: 953–955.
• 12.   Westenbrink BD, Voors AA, de Boer RA, Schuringa JJ, Klinkenberg T, van der Harst P, et al. Bone marrow dysfunction in chronic heart failure patients. Eur J Heart Fail 2010; 12: 676–684.
• 13.   Mojadidi MK, Galeas JN, Goodman-Meza D, Eshtehardi P, Msaouel P, Kelesidis I, et al. Thrombocytopaenia as a prognostic indicator in heart failure with reduced ejection fraction. Heart Lung Circ 2016; 25: 568–575.
• 14.   Ponikowski P, Voors AA, Anker SD, Bueno H, Cleland JGF, Coats AJS, et al. 2016 ESC Guidelines for the diagnosis and treatment of acute and chronic heart failure: The Task Force for the diagnosis and treatment of acute and chronic heart failure of the European Society of Cardiology (ESC). Developed with the special contribution of the Heart Failure Association (HFA) of the ESC. Eur Heart J 2016; 37: 2129–2200.
• 15.   McKee PA, Castelli WP, McNamara PM, Kannel WB. The natural history of congestive heart failure: The Framingham study. N Engl J Med 1971; 285: 1441–1446.
• 16.   Konishi M, Ishida J, Springer J, von Haehling S, Akashi YJ, Shimokawa H, et al. Heart failure epidemiology and novel treatments in Japan: Facts and numbers. ESC Heart Fail 2016; 3: 145–151.
• 17.   Peterson PN, Rumsfeld JS, Liang L, Albert NM, Hernandez AF, Peterson ED, et al. A validated risk score for in-hospital mortality in patients with heart failure from the American Heart Association get with the guidelines program. Circ Cardiovasc Qual Outcomes 2010; 3: 25–32.
• 18.   Shiraishi Y, Kohsaka S, Abe T, Mizuno A, Goda A, Izumi Y, et al. Validation of the Get With The Guideline-Heart Failure risk score in Japanese patients and the potential improvement of its discrimination ability by the inclusion of B-type natriuretic peptide level. Am Heart J 2016; 171: 33–39.
• 19.   Filippatos G, Rossi J, Lloyd-Jones DM, Stough WG, Ouyang J, Shin DD, et al. Prognostic value of blood urea nitrogen in patients hospitalized with worsening heart failure: Insights from the Acute and Chronic Therapeutic Impact of a Vasopressin Antagonist in Chronic Heart Failure (ACTIV in CHF) study. J Card Fail 2007; 13: 360–364.
• 20.   Kanda Y. Investigation of the freely available easy-to-use software “EZR” for medical statistics. Bone Marrow Transplant 2013; 48: 452–458.
• 21.   Kandis H, Ozhan H, Ordu S, Erden I, Caglar O, Basar C, et al. The prognostic value of mean platelet volume in decompensated heart failure. Emerg Med J 2011; 28: 575–578.
• 22.   Jacob J, Martín-Sánchez FJ, Gil V, Llorens P. The prognostic usefulness of platelets in acute heart failure. Emerg Med J 2011; 28: 1079.
• 23.   Weber M, Hamm C. Role of B-type natriuretic peptide (BNP) and NT-proBNP in clinical routine. Heart 2006; 92: 843–849.
• 24.   Mitchell O, Feldman DM, Diakow M, Sigal SH. The pathophysiology of thrombocytopenia in chronic liver disease. Hepat Med 2016; 8: 39–50.
• 25.   Lemmer A, VanWagner L, Ganger D. Congestive hepatopathy: Differentiating congestion from fibrosis: Congestive hepatopathy. Clin Liver Dis 2017; 10: 139–143.
• 26.   Kamath S, Blann AD, Lip GY. Platelet activation: Assessment and quantification. Eur Heart J 2001; 22: 1561–1571.
• 27.   Gasparyan AY, Ayvazyan L, Mikhailidis DP, Kitas GD. Mean platelet volume: A link between thrombosis and inflammation? Curr Pharm Des 2011; 17: 47–58.
• 28.   Thompson CB, Love DG, Quinn PG, Valeri CR. Platelet size does not correlate with platelet age. Blood 1983; 62: 487–494.
• 29.   Thompson CB. From precursor to product: How do megakaryocytes produce platelets? Prog Clin Biol Res 1986; 215: 361–371.
• 30.   Kuter DJ. The physiology of platelet production. Stem Cells 1996; 14(Suppl 1): 88–101.
• 31.   Schäfer A, Fraccarollo D, Hildemann S, Christ M, Eigenthaler M, Kobsar A, et al. Inhibition of platelet activation in congestive heart failure by aldosterone receptor antagonism and ACE inhibition. Thromb Haemost 2003; 89: 1024–1030.
• 32.   Hosten AO. BUN and creatinine. In: Walker HK, Hall WD, Hurst JW, editors. Clinical methods: The history, physical, and laboratory examinations. Boston: Butterworths, 2011.
• 33.   Bemelmans MH, Gouma DJ, Buurman WA. Influence of nephrectomy on tumor necrosis factor clearance in a murine model. J Immunol 1993; 150: 2007–2017.
• 34.   Bock JS, Gottlieb SS. Cardiorenal syndrome: New perspectives. Circulation 2010; 121: 2592–2600.
• 35.   Bleeker JS, Hogan WJ. Thrombocytosis: Diagnostic evaluation, thrombotic risk stratification, and risk-based management strategies. Thrombosis 2011; 2011: 536062.
• 36.   Maryanovich M, Zahalka AH, Pierce H, Pinho S, Nakahara F, Asada N, et al. Adrenergic nerve degeneration in bone marrow drives aging of the hematopoietic stem cell niche. Nat Med 2018; 24: 782–791.
• 37.   Kurmani S, Squire I. Acute heart failure: Definition, classification and epidemiology. Curr Heart Fail Rep 2017; 14: 385–392.