Correlation Between the J-MACS, HeartMate3, and J-HeartMate Risk Scores and Their Prognostic Impact in Japanese Patients With Heart Failure　― A Retrospective Study ―

Abstract

Background: In addition to the J-HeartMate Risk Score (J-HMRS) and HeartMate 3 Risk Score (HM3RS), the J-MACS Risk Score (J-MACS-RS) was developed to predict death after left ventricular assist device (LVAD) implantation in Japanese patients with heart failure (HF). However, the correlation between these scores, the characteristics of high-risk patients as per these scores, and the mortality stratification of these scores in HF patients regardless of LVAD implantation are still not fully understood.

Methods and Results: Hospitalized patients with HF who underwent echocardiography and right heart catheterization were included (n=269). Patients at low or medium risk per the J-HMRS or HM3RS and at high risk per the J-MACS-RS (LM_J-HMH_J-MACS and LM_HM3H_J-MACS, respectively) were compared with those at low or medium risk per both scores (LM_J-HMLM_J-MACS and LM_HM3LM_J-MACS, respectively). The J-MACS-RS was well associated with the J-HMRS (r=0.66) and HM3RS (r=0.65). Patients with LM_J-HMH_J-MACS were older and showed a higher prevalence of ischemic etiology and history of cardiac surgery than those with LM_J-HMLM_J-MACS. LM_J-HMH_J-MACS and LM_HM3H_J-MACS showed higher serum creatinine levels and central venous pressure-to-pulmonary artery wedge pressure ratios than LM_J-HMLM_J-MACS and LM_HM3LM_J-MACS, respectively. All scores stratified the 3-year mortality in patients with HF.

Conclusions: The J-MACS-RS correlated well with the J-HMRS and HM3RS. These scores may predict 3-year mortality, even in Japanese HF patients, regardless of LVAD implantation.

The left ventricular assist device (LVAD) is a well-established therapeutic modality for patients with advanced heart failure (HF).^1–3 Nevertheless, the benefits of a LVAD depends on the conditions of the recipient and the facility⁴ The HeartMate II Risk Score (HMIIRS), which is simply calculated using age, serum albumin (Alb) and serum creatinine (Cre) levels, the prothrombin time-international normalized ratio (PT-INR), and an index of the facilities’ experience of LVAD implantation, is a tool used to elucidate the risk associated with HeartMate II LVAD implantation.⁵ Recently, the HeartMate 3 Risk Score (HM3RS) was developed to stratify the mortality rate after HeartMate 3 LVAD implantation and its components include prior cardiac surgery, small LV, and a hemodynamic index representing right HF in addition to age and blood examination parameters such as blood urea nitrogen (BUN) and serum sodium (Na) levels.⁶ In Japan, the J-HeartMate Risk Score (J-HMRS), which is derived from the HMIIRS by modifying the facilities’ experience index according to the actual Japanese situation, is used as a selection criterion for destination therapy (DT) in patients with advanced HF.⁷

However, these risk scores were developed without including Japanese cohorts. Hence, the J-MACS risk score (J-MACS-RS) was established as a risk calculator for the 3-year mortality rate following LVAD implantation in Japanese patients. It has 4 components: age, history of cardiac surgery, Cre level, and hemodynamic index of right HF.⁸ Using J-MACS-RS, the 3-year mortality rate was 53%, 15%, and 6% in high-, medium-, and low-risk cohorts, respectively. Moreover, the area under the curve (AUC) was higher than those of the HMIIRS, J-HMRS, and HM3RS in Japanese patients.⁸ However, the J-MACS-RS and HM3RS are not always available for clinical application, because they require invasive hemodynamic evaluation. If the J-HMRS is well correlated with the J-MACS-RS, the J-HMRS is still useful for risk assessment of LVAD implantation in patients before undergoing right heart catheterization (RHC), or for those not indicated for RHC due to instability. Because the J-MACS-RS includes 2 different components from the J-HMRS (i.e., history of cardiac surgery and hemodynamic index), there is a possible discrepancy between the 2 risk scores.

Although patients at low or medium risk as per the J-HMRS but at high risk as per the J-MACS-RS are eligible for DT in the Japanese criteria, such patients are potentially at a high risk for LVAD implantation. There are no data regarding the association between the J-HMRS and J-MACS-RS. The HM3RS and J-MACS-RS have common factors such as history of cardiac surgery and a hemodynamic index of right HF, but there are no data regarding the relationship between them.⁹ Hence, this study aimed to elucidate the relationship between the J-HMRS, HM3RS, and J-MACS-RS for Japanese patients hospitalized with HF regardless of future LVAD implantation, particularly to identify the subset (high risk in the J-MACS-RS but low or medium risk in the J-HMRS or HM3RS) potentially at risk.

Moreover, these risk scores potentially reflect end-organ dysfunction and the severity of HF. Therefore, we hypothesized that these risk scores could estimate prognosis regardless of whether the patient underwent LVAD implantation. We also investigated the ability of the 3 risk scores to serve as prognostic tools in patients with HF.

Methods

Study Participants

Patients admitted to Chiba University Hospital between July 2014 and December 2022 with symptomatic HF who underwent both RHC and transthoracic echocardiography (TTE) during their hospital stay were retrospectively included. Patients with endstage renal failure requiring maintenance dialysis and those who had previously undergone LVAD implantation or heart transplantation were excluded; however, patients requiring temporary mechanical circulatory support devices and discharged alive were included.

Data Collection

All the patients underwent blood examinations, TTE, and RHC during hospitalization. TTE and RHC were performed after decongestion was completed. For patients requiring temporary mechanical circulatory devices, TTE and RHC were performed after device removal. Blood examinations were based on data obtained at discharge. The data were extracted from hospital records, along with demographic data and baseline characteristics.

Echocardiographic parameters were assessed according to the current American Society of Echocardiography guidelines.¹⁰ Metrics included LV end-diastolic diameter (LVDd), LV end-systolic diameter, LV ejection fraction (LVEF), tricuspid regurgitation pressure gradient, right ventricular (RV) end-diastolic diameter, and inferior vena cava diameter.

RHC assessed the mean central venous pressure (CVP), RV pressure, pulmonary artery pressure (PAP), mean pulmonary arterial wedge pressure (PAWP), and cardiac output (CO) using the Fick principle. Cardiac index (CI), pulmonary artery pulsatility index (PAPi), and RV stroke work index (RVSWI) were calculated using the following equations:

CI [L/min/m²] = CO [L/min] / body surface area (BSA)* [m²].

PAPi = (systolic PAP [mmHg] − diastolic PAP [mmHg]) / CVP [mmHg]¹¹

RVSWI [g × m/m²/beat] = 0.0136 × (mean PAP [mmHg] − CVP [mmHg]) × CO [L/min] / HR [beats/min] / BSA* [m²].¹²

BSA [m²] = (body weight [kg]^0.425) × (height [m]^0.725) × 0.007184.¹³

The J-HMRS, HM3-RS, and J-MACS-RS were calculated using the following formulas, and the low-, medium-, and high-risk patients were defined according to these scores.

(1) J-HMRS = 0.027 × age − 0.723 × Alb [g/dL] + 0.74 × Cre [mg/dL] + 1.136 × PT-INR** + 0.807 (if the facility’s experience of LVAD implantation is <3 per 2 years)

**PT-INR defined as 1 if the patient taking warfarin

Low risk<1.58≤medium risk≤2.48<high risk.^5,7

(2) HM3RS = 0.03496 × age + 0.53029 (if previous coronary artery bypass grafting or valve surgery) − 0.04112 × (Na [mEq/L] − 136) + 0.01093 × BUN [mg/dL] + 0.62149 (if LVDd <5.5 cm) + 0.44785 (if CVP/PAWP >0.6)

Low risk<2.41≤medium risk<2.97≤high risk⁶

(3) J-MACS-RS = 0.105 × age + 2.06 (if history of cardiac surgery) + 3.56 × Cre [mg/dL] + 2.61 (if CVP/PAWP >0.71)

Low risk<10.2≤medium risk<13.7≤high risk.⁸

Data Analysis

First, the distribution of patients as per the 3 risk scores was analyzed, and correlations between the J-MACS-RS and J-HMRS and between the J-MACS-RS and HM3RS were analyzed. The main objective of this study was to identify patients at potentially high risk as per the J-MACS-RS who could not be identified using other 2 risk scores. Patients classified as high risk with the J-MACS-RS and low or medium risk with the J-HMRS were defined as the LM_J-HMH_J-MACS group. Similarly, patients classified as high risk with the J-MACS-RS but low or medium risk with the HM3RS were defined as the LM_HM3H_J-MACS group. In contrast, low- or medium-risk patients with both the J-MACS-RS and J-HMRS and with both the J-MACS-RS and HM3RS were defined as the LM_J-HMLM_J-MACS and LM_HM3LM_J-MACS groups, respectively. The characteristics of the LM_J-HMH_J-MACS and LM_HM3H_J-MACS groups were analyzed and compared with those of the LM_J-HMLM_J-MACS and LM_HM3LM_J-MACS groups, respectively.

Second, the included patients were followed up for survival analysis for 3 years to examine whether the 3 risk scores stratified the mortality risk of patients with HF regardless of whether they underwent LVAD implantation. Patients surviving during the 3-year follow-up period with a durable LVAD newly implanted within the period were considered as survival.

Statistical Analysis

JMP Pro 17 statistical software (SAS Institute, Cary, NC, USA) was used for statistical analyses. Continuous variables are presented as mean±standard deviation, and Student’s t-test was used for group comparisons with normally distributed data. Continuous variables that were not normally distributed are presented as medians with interquartile ranges and were compared using the Mann-Whitney U test. Categorical variables are presented as proportions (%) and were analyzed using the chi-square test. Bivariate correlations were assessed using Pearson’s correlation coefficient (r). Statistical significance was set at P<0.05.

Receiver operating characteristic (ROC) analyses were performed, and the AUC for predicting 3-year mortality was calculated for the J-HMRS, HM3RS, and J-MACS-RS. Kaplan-Meier methods were used to evaluate survival, and log-rank testing was used to compare survival between each risk category in the 3 risk scores. The curves account for all patients, including those censored due to follow-up periods shorter than 3 years or death occurring during the observation period. Survival rates at 3 years were calculated after including these censored cases.

Ethics

This study was conducted in accordance with the Declaration of Helsinki and approved by the Ethics Committee of the Graduate School of Medicine, Chiba University (No. HK202409-01). The requirement for written informed consent was waived because of complete anonymization of the data and the observational study design. However, information about the study was made publicly available, and all research participants had the opportunity to refuse to participate.

Results

Patients’ Characteristics

The characteristics of the 269 patients included in this study are shown in Table 1. During the 3-year follow-up period, 14 patients underwent durable LVAD implantation.

Table 1.

Baseline Patient Characteristics

	n=269
Age [years]	58±16
Male, n (%)	206 (76.6)
ICM, n (%)	30 (10.7)
Median duration of HF [months] (IQR)	4 (1, 66)
BMI [kg/m2]	24.8±6.7
BNP [pg/mL]	241.0 (113.0, 497.0)
Cre [mg/dL]	1.1±0.5
Alb [g/dL]	3.7±0.6
Na [mmol/L]	138.1±3.4
PT-INR	1.2±0.5
Hypertension, n (%)	108 (40.1)
Diabetes mellitus, n (%)	88 (32.7)
Dyslipidemia, n (%)	108 (40.1)
Atrial fibrillation, n (%)	77 (28.6)
History of cardiac surgery, n (%)	32 (11.3)
Medication use
ACEi or ARB or ARNI	215 (79.9)
β-blocker	233 (86.6)
MRA	199 (74.0)
SGLT2i	66 (24.5)
Loop diuretic	231 (85.9)
TTE
LVEF [%]	31±15
LVDd [mm]	61±11
LVDs [mm]	53±13
TRPG [mmHg]	32±13
RVDd [mm]	41±8
IVC [mm]	18±6
Valvular disease	144 (53.5)
Moderate or severe AR	12 (4.5)
Moderate or severe MR	85 (31.6)
Moderate or severe TR	47 (17.5)
Right heart catheterization findings
CVP [mmHg]	8±5
mean PAP [mmHg]	27±11
PAWP [mmHg]	19±9
CVP/PAWP	0.47±0.24
CI (Fick) [L/min/m2]	2.2±0.8
PAPi	3.4±3.0
RVSWI	7.7±3.5

Unless indicated otherwise, data are given as the mean±standard deviation or n (%). ACEi, angiotensin-converting enzyme inhibitor; Alb, albumin; AR, aortic regurgitation; ARB, angiotensin-receptor blocker; ARNI, angiotensin-receptor neprilysin inhibitor; BMI, body mass index; BNP, B-type natriuretic peptide; CI, cardiac index; Cre, creatinine; CVP, central venous pressure; ICM, ischemic cardiomyopathy; HF, heart failure; IVC, inferior vena cava; LVDd, left ventricular end-diastolic diameter; LVDs, left ventricular end-systolic diameter; LVEF, left ventricular ejection fraction; PT-INR, prothrombin time-international normalized ratio; MRA, mineralocorticoid-receptor antagonist; SGLT2i, sodium-glucose cotransporter-2 inhibitor; TRPG, tricuspid regurgitation pressure gradient; RVDd, right ventricular end-diastolic diameter; MR, mitral regurgitation; PAP, pulmonary artery pressure; PAWP, pulmonary artery wedge pressure; PAPi, pulmonary artery pulsatility index; RVSWI, right ventricular stroke work index; TR, tricuspid regurgitation; TTE, transthoracic echocardiography.

Distribution of the 3 Risk Scores and Their Correlations

The J-HMRS and HM3RS correlated well with the J-MACS-RS (r=0.66, P<0.01 for the J-HMRS and J-MACS-RS; r=0.65, P<0.01 for the HM3RS and J-MACS-RS) (Figure 1).

Figure 1.

Correlation between the J-MACS-RS and J-HMRS and between the J-MACS-RS and HM3RS. J-HMRS: Low risk<1.58≤medium risk≤2.48<high risk; HM3RS: Low risk<2.41≤medium risk<2.97≤high risk; J-MACS-RS: Low risk<10.2≤medium risk<13.7≤high risk. Patients classified as high risk with the J-MACS-RS despite having low- or medium-risk J-HMRS were defined as the LM_J-HMH_J-MACS group. Patients classified as at high risk with the HM3RS but as low or medium risk with the HM3RS were defined as the LM_HM3H_J-MACS group. Low- or medium-risk patients with both the J-MACS-RS and J-HMRS and both the J-MACS-RS and HM3RS were defined as the LM_J-HMLM_J-MACS and LM_HM3LM_J-MACS groups, respectively. J-HMRS, J-HeartMate Risk Score; HM3RS, HeartMate3 Risk Score; J-MACS-RS, J-MACS Risk Score.

The distribution of each risk score is shown in Figure 2. The distribution of high risk in each risk score was: J-MACS-RS, 17.8%; J-HMR, 6.7%; and HM3RS, 20%. A total of 37 patients (13.7% of the entire cohort) were assigned to the LM_J-HMH_J-MACS group and 19 patients (7.1%) to the LM_HM3H_J-MACS group.

Figure 2.

Distribution by respective risk in the J-MACS-RS and J-HMRS and in the J-MACS-RS and HM3RS. Some patients exhibit two divergent risk scores. The LM_J-HMH_J-MACS, LM_HM3H_J-MACS, LM_J-HMLM_J-MACS, and LM_HM3LM_J-MACS groups are shown. Patients classified as high risk with the J-MACS-RS despite having low or medium risk with the J-HMRS were defined as the LM_J-HMH_J-MACS group. Patients classified as high risk in the HM3RS group but as low or medium risk in the HM3RS group were defined as the LM_HM3H_J-MACS group. Low- or medium-risk patients with both the J-MACS-RS and J-HMRS and both the J-MACS-RS and HM3RS were defined as the LM_J-HMLM_J-MACS and LM_HM3LM_J-MACS groups, respectively. J-HMRS, J-HeartMate risk score; HM3RS, HeartMate3 risk score; J-MACS-RS, J-MACS risk score.

Comparison Between J-HMRS and J-MACS-RS

The comparison of the LM_J-HMH_J-MACS and LM_J-HMLM_J-MACS groups is shown in Table 2. There were no significant differences in sex or disease duration between groups. The LM_J-HMH_J-MACS group was significantly older and presented with a higher Cre level and a higher prevalence of ischemic etiology, hypertension, diabetes, and a history of cardiac surgery than the LM_J-HMLM_J-MACS group. This group also had a higher LVEF and incidence of CVP/PAWP >0.71. Additionally, there were some differences between the groups in the pharmacological treatments. The proportion of patients on angiotensin-converting enzyme inhibitors (ACEi), angiotensin II receptor blockers (ARB), angiotensin-receptor-neprilysin inhibitors (ARNI), and mineralocorticoid-receptor antagonist (MRA) was significantly lower in the LM_J-HMH_J-MACS group than in the LM_J-HMLM_J-MACS group.

Table 2.

Comparison Between the LM_J–HMH_J–MACS and LM_J–HMLM_J–MACS Groups

	LMJ–HMHJ–MACS	LMJ–HMLMJ–MACS	P value
Age [years]	68±14	56±16	<0.01
Male, n (%)	29 (78)	162 (75)	0.69
ICM, n (%)	7 (19)	17 (8)	0.04
Median duration of HF [months] (IQR)	20 (2, 76)	3 (1, 37)	0.18
BMI [kg/m2]	25.2±6.4	25.0±7.0	0.87
BNP [pg/mL] (IQR)	262 (155, 680)	221 (101, 464)	0.45
Cre [mg/dL]	1.6±0.5	1.0±0.3	<0.01
Alb [g/dL]	3.6±0.5	3.8±0.5	0.08
Na [mmol/L]	137.8±3.8	138.1±3.2	0.51
PT-INR	1.1±0.3	1.1±0.3	0.73
Hypertension, n (%)	21 (57)	77 (38)	0.02
Diabetes mellitus, n (%)	17 (46)	63 (29)	0.04
Dyslipidemia, n (%)	16 (43)	82 (40)	0.56
Atrial fibrillation, n (%)	13 (35)	59 (28)	0.35
History of cardiac surgery, n (%)	10 (27)	12 (6)	<0.01
Medication use, n (%)
ACEi or ARB or ARNI	25 (68)	181 (85)	0.01
β-blocker	28 (76)	188 (88)	0.05
MRA	20 (54)	166 (78)	<0.01
SGLT2i	13 (35)	46 (21)	0.07
Loop diuretic	32 (86)	185 (86)	1.00
TTE
LVEF [%]	37±17	30±14	0.01
LVDd [mm]	60±11	61±11	0.73
LVDs [mm]	50±14	53±13	0.25
TRPG [mmHg]	32±16	32±13	0.98
RVDd [mm]	43±10	40±8	0.07
IVC [mm]	19±7	17±5	0.26
Valve disease, n (%)
Moderate or severe AR	3 (8)	9 (4)	0.30
Moderate or severe MR	8 (22)	75 (35)	0.11
Moderate or severe TR	11 (30)	33 (15)	0.04
Right heart catheterization
CVP [mmHg]	10±6	8±4	0.02
PAP (m) [mmHg]	26±11	27±10	0.67
PAWP [mmHg]	16±7	18±9	0.12
CVP/PAWP	0.63±0.26	0.44±0.21	<0.01
CVP/PAWP >0.71, n (%)	17 (46)	22 (10)	<0.01
CI (Fick) [L/min/m2]	2.2±0.6	2.1±0.7	0.71
PAPi	3.1±3.0	3.4±3.0	0.53
RVSWI	8.2±5.3	7.6±3.2	0.55

Unless indicated otherwise, data are given as the mean±SD or n (%). Abbreviations as in Table 1.

Comparison Between HM3RS and J-MACS-RS

The comparison of the LM_HM3H_J-MACS and LM_HM3LM_J-MACS groups is presented in Table 3. The LM_HM3H_J-MACS group presented a higher prevalence of diabetes and dyslipidemia, higher Cre level, and lower Alb level than the LM_HM3LM_J-MACS group. There were no significant differences in the echocardiographic parameters. Regarding hemodynamic parameters, the CVP/PAWP value and the proportion of CVP/PAWP >0.71 were significantly higher in the LM_HM3H_J-MACS group than in the LM_HM3LM_J-MACS group. Additionally, there were some differences between the groups in the pharmacological treatments. The proportion of patients on MRAs was significantly lower in the LM_HM3H_J-MACS than in the LM_HM3LM_J-MACS group.

Table 3.

Comparison Between the LM_HM3H_J–MACS and LM_HM3LM_J–MACS Groups

	LMHM3HJ-MACS	LMHM3LMJ-MACS	P value
Age [years]	58±15	53±14	0.22
Male, n (%)	17 (89)	153 (78)	0.24
ICM, n (%)	2 (11)	16 (8)	0.72
Median duration of HF [months] (IQR)	3 (1, 78)	3 (1, 25)	0.37
BMI [kg/m2]	27.1±7.7	25.4±7.2	0.37
BNP [pg/mL] (IQR)	487 (234, 860)	220 (100, 463)	0.05
Cre [mg/dL]	2.1±0.9	1.0±0.3	<0.01
Alb [g/dL]	3.5±0.5	3.8±0.5	0.02
Na [mmol/L]	137.6±3.1	138.3±3.1	0.36
PT-INR	1.1±0.2	1.1±0.3	0.26
Hypertension, n (%)	10 (53)	71 (36)	0.16
Diabetes mellitus, n (%)	10 (53)	59 (30)	0.04
Dyslipidemia, n (%)	12 (63)	75 (38)	0.04
Atrial fibrillation, n (%)	5 (26)	51 (26)	0.98
History of cardiac surgery, n (%)	1 (5)	5 (3)	0.49
Medication use
ACEi or ARB or ARNI	17 (89)	169 (86)	0.69
β-blocker	15 (79)	178 (91)	0.10
MRA	8 (42)	156 (80)	<0.01
SGLT2i	6 (32)	39 (20)	0.23
Loop diuretic	17 (89)	173 (88)	0.88
TTE
LVEF [%]	29±11	28±12	0.80
LVDd [mm]	64±8	62±11	0.42
LVDs [mm]	56±10	55±12	0.55
TRPG [mmHg]	30±13	31±13	0.74
RVDd [mm]	41±10	40±7	0.71
IVC [mm]	17±7	18±5	0.48
Valve disease, n (%)
Moderate or severe AR	2 (11)	8 (4)	0.20
Moderate or severe MR	5 (26)	65 (33)	0.60
Moderate or severe TR	3 (16)	26 (13)	0.76
Right heart catheterization
CVP [mmHg]	10±8	8±4	0.13
PAP (m) [mmHg]	25±13	27±10	0.34
PAWP [mmHg]	16±8	19±9	0.20
CVP/PAWP	0.6±0.3	0.4±0.2	0.01
CVP/PAWP >0.6, n (%)	10 (42)	38 (19)	<0.01
CVP/PAWP >0.71, n (%)	7 (37)	21 (11)	<0.01
CI (Fick) [L/min/m2]	2.2±0.9	2.1±0.7	0.45
PAPi	3.2±4.0	3.5±3.2	0.78
RVSWI	6.8±3.7	7.6±3.2	0.34

Unless indicated otherwise, data are given as the mean±SD or n (%). Abbreviations as in Table 1.

Survival Analysis

The Kaplan-Meier analysis included 269 patients, of whom 38 (17%) died during the follow-up period (median follow-up period: 1,095 days).

Regarding the J-HMRS, 15 deaths (7.5%) had occurred among the 201 patients in the low-risk group, 18 deaths (36.0%) among 50 patients in the medium-risk group, and 5 deaths (27.8%) among 18 high-risk patients at the 3-year follow-up. Regarding the HM3RS, 9 deaths (6.4%) had occurred among 141 low-risk patients, 12 deaths (16.2%) among 74 medium-risk patients, and 17 deaths (31.5%) among 54 high-risk patients at the 3-year follow-up. Regarding the J-MACS-RS, a total of 4 deaths (4.0%) had occurred among 99 low-risk patients, 25 deaths (20.5%) among 122 medium-risk patients, and 9 deaths (18.8%) among 48 high-risk patients at 3-year follow-up. The survival curves for each risk score are shown in Figure 3. For all scores, there were significant differences among the risk categories (P<0.01). The ROC curves are shown in Figure 4. The AUCs for the J-HMRS, HM3RS, and J-MACS-RS were 0.78, 0.73, and 0.68, respectively.

Figure 3.

Survival curve (median 1,095 days) by risk in the J-MACS-RS, HM3RS, and J-HMRS. J-HMRS, J-HeartMate Risk Score; HM3RS, HeartMate3 Risk Score; J-MACS-RS, J-MACS Risk Score.

Figure 4.

ROC curves of J-MACS-RS, HM3RS, and J-HMRS at 3-year follow-up. ROC, receiver operating characteristic; J-HMRS, J-HeartMate Risk Score; HM3RS, HeartMate3 Risk Score; J-MACS-RS, J-MACS Risk Score.

Discussion

The present study examined the associations between the J-HMRS, HM3RS, and J-MACS-RS among Japanese patients with HF and identified the characteristics of patients at a potentially high risk of LVAD implantation by comparing the 3 risk scores. Furthermore, we validated the prognostic usefulness of these risk scores in patients with HF without LVAD implantation. The main findings of this study were: (1) J-HMRS and HM3RS correlated well with the J-MACS-RS; (2) the LM_J-HMH_J-MACS group was older, had higher Cre levels, and a higher prevalence of ischemic etiology and cardiac surgery with some vascular risk factors than the LM_J-HMLM_J-MACS group; (3) the LM_J-HMH_J-MACS and LM_HM3H_J-MACS groups both presented with a higher prevalence of an elevated CVP/PAWP ratio compared with the LM_J-HMLM_J-MACS and LM_HM3LM_J-MACS groups; and (4) all the risk scores stratified the 3-year mortality of Japanese patients with HF who did not undergo LVAD implantation. Among them, the J-HMRS showed the highest AUC. To the best of our knowledge, this is the first study to directly analyze the correlation between the J-HMRS, HM3RS, and J-MACS-RS in patients with HF who did not undergo LVAD implantation.

Associations Among the 3 Risk Scores

It was predictable that the J-HMRS and HM3RS showed good correlation with the J-MACS-RS because the formulas for these risk scores include age and an index of renal function (BUN or Cre). Nevertheless, 37 patients were discordantly categorized into higher-risk subsets in the J-MACS-RS compared with the J-HMRS and were included in the LM_J-HMH_J-MACS group. The LM_J-HMH_J-MACS group presented with higher age and Cre levels than the LM_J-HMLM_J-MACS group, which may be due to the larger coefficients of age and Cre in the formula. However, the absolute values and ranges of the scores are different from each other, and the true impact of these coefficients on the scores is undetermined. In addition, the LM_J-HMH_J-MACS group presented with higher prevalence of hypertension and diabetes. These vascular risk factors can lead to ischemic heart disease and renal dysfunction. Some clinical investigations and systematic reviews have shown that ischemic etiology has less effect on overall survival;^14,15 however, bleeding complications are suggested to be elevated in patients with ischemic cardiomyopathy¹⁶ or a higher average brachial-ankle pulse wave velocity.¹⁷ Arteriosclerotic and ischemic heart diseases may affect the clinical course after LVAD implantation.

The HM3RS and J-MACS-RS formulas have many components in common: age, history of cardiac surgery, indices of renal function (BUN or Cre), and CVP/PAWP elevation. Therefore, the number of patients with a discrepancy in the risk category between the HM3RS and J-MACS-RS was small. Among them, an extremely high Cre level and CVP/PAWP >0.71 were associated with a relatively higher J-MACS-RS compared with the HM3RS. The cutoff value of CVP/PAWP for point addition is different between these risk scores (>0.71 for the J-MACS-RS and >0.60 for the HM3RS). Therefore, the higher cutoff value of the J-MACS-RS was associated with the LM_HM3H_J-MACS group. Because right heart function affects the outcome of LVAD implantation,^18–20 these risk scores emphasize the CVP/PAWP ratio rather than LV function. Notably, LVEF was not associated with the LM_J-HMH_J-MACS and LM_HM3H_J-MACS groups. In summary, patients with extremely advanced renal dysfunction and right HF hemodynamics should be carefully determined for LVAD eligibility if the J-HMRS shows low or medium risk, because the J-MACS-RS tends to be relatively elevated compared with the J-HMRS and HM3RS.

Risk Scores as Prognostic Tools

We hypothesized that these risk scores would be useful for predicting HF prognosis in patients, regardless of whether they undergo LVAD implantation, because the formulas contain indices related to nutrition, end-organ dysfunction, and right HF in addition to age. All 3 scores successfully stratified the 3year mortality of patients with HF regardless of LVAD treatment, suggesting that they are useful for estimating mortality risk in patients with HF. For each risk score, the 3-year mortality rates with low risk were <10%. However, the Kaplan-Meier curves showed different trends in medium and high risk for the 3 scores, which largely depends on the cutoff value. One implication of this analysis is that the prognosis of low-risk HF patients is good in all 3 risk scores, regardless of whether they undergo LVAD implantation. In particular, the AUC was the highest for the J-HMRS, followed by the HM3RS and J-MACS-RS. The HM3RS and J-MACS-RS place importance on a history of cardiac surgery and right HF hemodynamics as risk factors after LVAD implantation; however, the prognostic impact of previous cardiac surgery on HF treatment is relatively small. Right HF is a major risk factor after LVAD implantation; however, its impact on the prognosis of HF patients regardless of whether they undergo LVAD treatment may not be definitive because LV systolic and diastolic functions are also important.

There are several risk scores that can be used to predict the prognosis of patients with HF. The Seattle Heart Failure Model (SHFM) is a representative model.²¹ Lanfear et al. validated the accuracy of the SHFM in patients with non-inotrope-dependent advanced HF using the Risk Assessment and Comparative Effectiveness of Left Ventricular Assist Device and Medical Management in Ambulatory Heart Failure Patients (ROADMAP) study.²² The AUC for survival of the SHFM was 0.71, suggesting that the score is a significant predictor of death among patients with advanced HF and LVAD candidates. The AUC of the present study was comparable to that for the SHFM, even though the characteristics of the patients differed.

The J-HMRS is simple and easy to calculate but appropriately reflects nutritional status and renal and hepatic function with 3 biomarkers, which may also reflect the conditions of congestion and hypoperfusion. Yoshimura et al. investigated the prognosis of 198 older non-responders to cardiac resynchronization therapy using the J-HMRS.²³ They reported that the mortality rate was higher in the high-risk cohort of the J-HMRS than in the low-risk cohort. Overall, the J-HMRS can be used as a mortality predictor in patients with HF patients regardless of whether they undergo LVAD implantation.

Study Limitations

First, this was a retrospective study, and there were potential selection biases. In particular, the patients included a certain number of “stage C” HF patients hospitalized for the first time due to HF onset and who underwent RHC as the initial pathophysiological assessment. Generally, only “stage D” HF patients require LVAD implantation; however, the present study did not include only such patients, because the main purpose of this study was to analyze the association between 3 risk scores. Therefore, it requires attention in the interpretation of the results. For example, the mortality rate was probably lower than that of real LVAD candidates. Prospective studies with rigorous criteria for selecting patients with advanced HF are required. Second, we performed this study based on the hypothesis that patients in the LM_J-HMH_J-MACS group were potentially at high risk for LVAD implantation; however, this was not validated. The risk of LVAD implantation for patients with high risk on the J-MACS-RS, regardless of the J-HMRS, should be validated with an appropriate cohort of LVAD recipients in a future investigation.

Conclusions

The present study results suggested that the J-MACS-RS correlates well with the J-HMRS and HM3RS. Patients with extremely advanced renal dysfunction and right HF hemodynamics should be carefully determined for LVAD eligibility if the J-HMRS shows low or medium risk, because the J-MACS-RS tends to be relatively elevated compared with the J-HMRS. The t3 risk scores can be used to predict the 3-year mortality, even in Japanese patients with HF, regardless of whether they undergo LVAD implantation.

Acknowledgments

We thank Editage (www.editage.jp) for the English language editing.

Disclosures

Conflicts of Interest: Y.K. is a member of Circulation Journal’s Editorial Team and received research grants from Abbott Medical Japan and Nipro. The remaining authors have no conflicts of interest to disclose. Sources of Funding: None.

IRB Information

This study was approved by the Ethics Committee of the Graduate School of Medicine, Chiba University (No. HK202409-01) and was conducted in accordance with the principles of the Declaration of Helsinki.

Data Availability

The deidentified participant data will not be shared.

References

• 1.   Rose EA, Gelijns AC, Moskowitz AJ, Heitjan DF, Stevenson LW, Dembitsky W, et al. Long-term use of a left ventricular assist device for end-stage heart failure. N Engl J Med 2001; 345: 1435–1443.
• 2.   Goldstein DJ, Naka Y, Horstmanshof D, Ravichandran AK, Schroder J, Ransom J, et al. Association of clinical outcomes with left ventricular assist device use by bridge to transplant or destination therapy intent: The multicenter study of MagLev technology in patients undergoing mechanical circulatory support therapy with HeartMate 3 (MOMENTUM 3) randomized clinical trial. JAMA Cardiol 2020; 5: 411–419.
• 3.   Sidhu K, Lam PH, Mehra MR. Evolving trends in mechanical circulatory support: Clinical development of a fully magnetically levitated durable ventricular assist device. Trends Cardiovasc Med 2020; 30: 223–229.
• 4.   Yoshioka D, Toda K, Ono M, Nakatani T, Shiose A, Matsui Y, et al. Clinical results, adverse events, and change in end-organ function in elderly patients with HeartMateII left ventricular assist device: Japanese multicenter study. Circ J 2018; 82: 409–418.
• 5.   Cowger J, Sundareswaran K, Rogers JG, Park SJ, Pagani FD, Bhat G, et al. Predicting survival in patients receiving continuous flow left ventricular assist devices: The HeartMate II risk score. J Am Coll Cardiol 2013; 61: 313–321.
• 6.   Mehra MR, Nayak A, Morris AA, Lanfear DE, Nemeh H, Desai S, et al. Prediction of survival after implantation of a fully magnetically levitated left ventricular assist device. JACC Heart Fail 2022; 10: 948–959.
• 7.   Kinugawa K, Sakata Y, Ono M, Nunoda S, Toda K, Fukushima N, et al. Consensus Report on Destination Therapy in Japan: From the DT Committee of the Council for Clinical Use of Ventricular Assist Device Related Academic Societies. Circ J 2021; 85: 1906–1917.
• 8.   Imamura T, Kinugawa K, Nishimura T, Toda K, Saiki Y, Niinami H, et al. Novel scoring system to risk stratify patients receiving durable left ventricular assist device from J-MACS Registry data. Circ J 2023; 87: 1103–1111.
• 9.   Kametani M, Minami Y, Hattori H, Haruki S, Yamaguchi J. Relationship between the HeartMate Risk Score category on admission and outcome in patients with acute heart failure referred to a cardiac intensive care unit. Heart Vessels 2025; 40: 55–61.
• 10.   Lang RM, Badano LP, Mor-Avi V, Afilalo J, Armstrong A, Ernande L, et al. Recommendations for cardiac chamber quantification by echocardiography in adults: An update from the American Society of Echocardiography and the European Association of Cardiovascular Imaging. Eur Heart J Cardiovasc Imaging 2015; 16: 233–270.
• 11.   Lim HS, Gustafsson F. Pulmonary artery pulsatility index: Physiological basis and clinical application. Eur J Heart Fail 2020; 22: 32–38.
• 12.   Jumatate R, Ingvarsson A, Smith GJ, Roijer A, Ostenfeld E, Waktare J, et al. Right ventricular stroke work index by echocardiography in adult patients with pulmonary arterial hypertension. BMC Cardiovasc Disord 2021; 21: 219.
• 13.   Du Bois D, Du Bois EF. Clinical calorimetry: Tenth paper a formula to estimate the approximate surface area if height and weight be known. Arch Intern Med 1916; 17: 863–871.
• 14.   Tsiouris A, Borgi J, Karam J, Nemeh HW, Paone G, Brewer RJ, et al. Ischemic versus nonischemic dilated cardiomyopathy: The implications of heart failure etiology on left ventricular assist device outcomes. ASAIO J 2013; 59: 130–135.
• 15.   Wavell C, Sokolowski A, Klingel ML, Yin C, Nagpal AD. Clinical effectiveness of therapy with continuous-flow left ventricular assist devices in nonischemic versus ischemic cardiomyopathy: A systematic review and meta-analysis. Can J Surg 2021; 64: e39–e47.
• 16.   Chou BP, Critsinelis A, Lamba HK, Long G, Civitello AB, Delgado RM, et al. Continuous-flow left ventricular assist device support in patients with ischemic versus nonischemic cardiomyopathy. Tex Heart Inst J 2021; 48: e207241.
• 17.   Imaoka S, Yoshioka D, Saito S, Kawamura T, Kawamura A, Toda K, et al. Clinical outcomes of left ventricular assist device bleeding complication. ASAIO J 2025; 71: 229–234.
• 18.   Kapelios CJ, Lund LH, Wever-Pinzon O, Selzman CH, Myers SL, Cantor RS, et al. Right heart failure following left ventricular device implantation: Natural history, risk factors, and outcomes Circ Heart Fail 2022; 15: e008706.
• 19.   Kormos RL, Teuteberg JJ, Pagani FD, Russell SD, John R, Miller LW, et al. Right ventricular failure in patients with the HeartMate II continuous-flow left ventricular assist device: Incidence, risk factors, and effect on outcomes. J Thorac Cardiovasc Surg 2010; 139: 1316–1324.
• 20.   Rich JD, Gosev I, Patel CB, Joseph S, Katz JN, Eckman PM, et al. The incidence, risk factors, and outcomes associated with late right-sided heart failure in patients supported with an axial-flow left ventricular assist device. J Heart Lung Transplant 2017; 36: 50–58.
• 21.   Levy WC, Mozaffarian D, Linker DT, Sutradhar SC, Anker SD, Cropp AB, et al. The Seattle Heart Failure Model: Prediction of survival in heart failure. Circulation 2006; 113: 1424–1433.
• 22.   Lanfear DE, Levy WC, Stehlik J, Estep JD, Rogers JG, Shah KB, et al. Accuracy of Seattle Heart Failure Model and HeartMate II Risk Score in non-inotrope-dependent advanced heart failure patients: Insights from the ROADMAP study (Risk assessment and comparative effectiveness of left ventricular assist device and medical management in ambulatory heart failure patients). Circ Heart Fail 2017; 10: e003745.
• 23.   Yoshimura A, Kikuchi N, Saito S, Suzuki A, Hattori H, Shoda M, et al. Stratification of destination therapy candidates by J-HeartMate Risk Score among elderly non-responders to cardiac resynchronization therapy. Circ Rep 2022; 4: 405–411.