5-Year Outcomes of Magnetically Levitated Left Ventricular Assist Device in Advanced Heart Failure　― Japanese Cohort ―

Abstract

Background: HeartMate 3 (HM3), a magnetically levitated centrifugal-flow pump, has demonstrated superior hemocompatibility and reduced adverse events compared to HeartMate II (HMII), an axial-flow pump, in global studies. However, because long-term comparative data in Japanese patients remain scarce, in the present study we evaluated the 5-year outcomes of HM3 support by comparing them with those of HMII at 2 leading left ventricular assist device (LVAD) centers in Japan.

Methods and Results: We retrospectively analyzed 364 patients who underwent primary LVAD implantation (HM3: n=168; HMII: n=196) between 2010 and 2023. The primary endpoint included survival to transplant, recovery, or continued LVAD support free from stroke or pump replacement. At 5 years, freedom from the composite endpoint was higher in the HM3 group (75% vs. 52%; hazard ratio [HR] 0.52; P=0.001), although overall survival was comparable (90% vs. 85%; P=0.44). The HM3 group experienced significantly fewer strokes (HR 0.40; P=0.0008), bleeding events (HR 0.22; P<0.0001), and pump thrombosis (HR 0.09; P=0.003). Rates of rehospitalization, driveline infections, and late right heart failure did not differ between the groups.

Conclusions: HM3 support significantly improved long-term event-free outcomes compared to HMII, despite comparable overall survival, supporting the use of HM3 as durable mechanical circulatory support devices in Japan.

Central Figure

Advanced heart failure is a major cause of death worldwide, and its prevalence continues to increase.¹ For patients with endstage disease, surgical interventions such as heart transplantation or left ventricular assist device (LVAD) implantation are often required.^2,3 In particular, LVAD therapy has gained prominence in recent years as a bridge-to-transplant (BTT) or destination therapy (DT),⁴ and in Japan, because transplant waiting times exceed 5 years, even patients implanted as BTT require prolonged LVAD support, highlighting the importance of real-world long-term outcome data.⁵

Editorial p 183

The HeartMate II (HMII; Abbott Laboratories, Chicago, IL, USA), an axial-flow pump, has been extensively used in clinical practice, but with several complications, including stroke, pump thrombosis, gastrointestinal bleeding, and mechanical failure.^6,7 In contrast, the HeartMate 3 (HM3; Abbott Laboratories) uses a fully magnetically levitated centrifugal pump, which minimizes friction at blood-contact surfaces, thereby improving hemocompatibility and reducing the risk of thrombosis, bleeding, and mechanical failure.^8–18 The INTERMACS registry¹⁸ and MOMENTUM 3 trial¹² have demonstrated that HM3 offers significant advantages over HMII in key clinical outcomes, including improved survival and reduced rates of pump thrombosis, stroke, and bleeding.

We previously reported favorable short-term outcomes of HM3 implantation at 2 high-volume centers in Japan, particularly regarding reduced rates of stroke and pump thrombosis.¹⁹ However, long-term comparative data for 5-year outcomes between HM3 and HMII in the Japanese population remain scarce. Owing to Japan’s prolonged transplant waiting times, even transplant-eligible younger patients require long-term LVAD support, offering evidence not reflected in international trials. Therefore, directly comparing HM3 with its predecessor HMII in Japan is warranted to confirm the appropriateness of this transition in real-world practice. Our aim was to compare the 5-year outcomes of HM3 support with those of HMII at 2 leading LVAD centers in Japan.

Methods

Study Design and Patient Selection

This retrospective observational study was conducted at The University of Osaka Hospital and the National Cerebral and Cardiovascular Center. Consecutive patients who underwent primary implantation of either HMII or HM3 between May 2010 and December 2023 were included. In Japan, HM3 received regulatory approval in 2019, after which all new HeartMate implantations were performed exclusively using this device. HMII was used between May 2010 and May 2019, whereas HM3 has been used for all cases since June 2019. Patients who underwent pump exchange from a previously implanted durable LVAD were excluded (Supplementary Figure 1).

A total of 168 patients with HM3 and 196 with HMII were included in the analysis. Of them, 104 patients with HM3 and 119 patients with HMII underwent implantation at the National Cerebral and Cardiovascular Center, while 64 patients with HM3 and 77 patients with HMII were treated at the University of Osaka Hospital. Implantation strategies were classified as BTT, bridge-to-candidacy, or DT. DT implantations with HMII were performed as part of a clinical trial in 2016 and 2017, whereas DT implantations with HM3 began after its official approval in May 2021. The primary endpoint was a composite of survival to transplant, recovery, or LVAD support free from stroke or pump replacement, defined in accordance with the framework of the MOMENTUM 3 trial.¹² Overall survival was also evaluated as a co-primary endpoint. Secondary endpoints included the incidences of stroke, bleeding, driveline infection, pump replacement, late right heart failure (RHF), and unplanned rehospitalization due to LVAD-related events. Data were extracted from the medical records of each institution.

This study was approved by the Institutional Review Board of Osaka University Hospital (approval number: 21372(T1), approval date: January 18, 2022), and written informed consent for publication was given by all patients. The study was conducted in accordance with the principles of the Declaration of Helsinki and the ethical standards of the institutional and national research committees.

Device Selection, Implantation Techniques, and Postoperative Management

Until 2022, the selection of durable LVAD devices, including HMII, HM3, HeartWare, DuraHeart, EVAHEART, and Jarvik, was determined at the discretion of each center through multidisciplinary team conferences. However, since 2022, HM3 has been used exclusively for all new implantations. Most procedures were performed via median sternotomy, with the inflow cannula inserted into the left ventricular apex and outflow graft anastomosed to the ascending aorta. Concomitant procedures, such as aortic or tricuspid valve surgeries, were performed at the discretion of the operating surgeon. Aortic valve interventions included valve replacement, patch closure, or Park’s stitch technique.

Postoperatively, all patients received anticoagulation with aspirin (100 mg/day) and warfarin, with a prothrombin time-international normalized ratio (PT-INR) target of 2.5 for HMII and 2.0 for HM3. Guideline-directed medical therapy was introduced and uptitrated whenever clinically tolerated after LVAD implantation at both participating centers.

Follow-up and Data Collection

Following discharge, patients were monitored monthly via laboratory testing, driveline site care, and imaging assessments. The median follow-up duration was 3.2 [2.0–4.4] years for HM3 and 4.1 [2.6–5.3] years for HMII.

Baseline data included demographics, medical history, laboratory data, echocardiographic parameters, and hemodynamic data from right heart catheterization. Bridge-to-bridge (BTB) was defined as the transition from temporary mechanical circulatory support, such as extracorporeal VAD or Impella® (Abiomed, Danvers, MA, USA) to durable LVAD. A history of prior cardiac surgery, including temporary VAD support, was recorded as a previous reoperation. Intraoperative data included total operative time, cardiopulmonary bypass (CPB) time, transfusion volume, and concomitant cardiac procedures. Postoperative echocardiography and right heart catheterization were performed after hemodynamic stabilization during the initial hospitalization.

Major adverse events during hospitalization and post-discharge LVAD support were reviewed. Stroke was defined as radiologically confirmed infarction or intracranial hemorrhage; transient ischemic attack or seizures without focal lesions on neuroimaging were not classified as stroke. Pump replacement was classified by etiology, including infection, thrombosis, or driveline fracture. Bleeding events were defined as any non-cerebral hemorrhage that required hospitalization. Driveline infection was defined as an exit-site infection requiring antibiotics or surgical debridement. Late RHF was defined as a condition requiring hospitalization in the chronic phase for intravenous inotropes, or adjustment of diuretics or pulmonary vasodilators. These definitions were based on or adapted from the International Society for Heart and Lung Transplantation consensus criteria for adverse event reporting in mechanical circulatory support. All LVAD-related rehospitalizations were unplanned; scheduled admissions were excluded. Patients who required right ventricular assist device (RVAD) support during the immediate postoperative period were not considered to have developed late RHF.

Statistical Analysis

Continuous variables are presented as medians and interquartile ranges and were compared using the Mann-Whitney U test. Categorical variables are presented as counts and percentages and were compared using the χ² test. Kaplan-Meier analysis was used to generate survival and event-free curves, with comparisons made using the log-rank test. A P value <0.05 was considered statistically significant. Prognostic factors for 5-year death and the composite outcome were assessed using univariate and multivariate Cox proportional hazards models. Statistical analyses were performed using JMP 17 software (SAS Institute Inc., Cary, NC, USA).

Results

Baseline Clinical Characteristics

Preoperative characteristics are summarized in Table 1. No significant differences were observed between the groups in demographic variables such as age, sex, and body surface area. The proportion of patients with dilated cardiomyopathy was lower in the HM3 group than in the HMII group. Regarding the clinical setting at the time of LVAD implantation, BTB implantation was more common in the HM3 group. Specifically, 19 patients (11%) in this group were bridged from Impella support, whereas the proportion bridged from extracorporeal VAD was comparable between groups (16% vs. 17%, P=0.59). In the HM3 group, 26 patients (15%) received the device as DT. No significant differences were observed in preoperative laboratory parameters, including hemoglobin, albumin, estimated glomerular filtration rate, PT–INR, and B-type natriuretic peptide. Although the left ventricular end-diastolic dimension (LVDd) was smaller in the HM3 group, other echocardiographic and hemodynamic findings were comparable between the groups. Regarding baseline valvular disease, mild aortic regurgitation was more frequent in the HM3 group than in the HMII group (35% vs. 18%, P=0.03). In contrast, the prevalence of tricuspid regurgitation was similar between groups. Baseline hemodynamic parameters, including pulmonary artery pulsatility index, right ventricular stroke work index, and pulmonary vascular resistance, were comparable between groups.

Table 1.

Baseline Clinical Characteristics

Variables	HeartMate 3 (n=168)	HeartMate II (n=196)	P value
Age, years	50 [36–58]	46 [36–56]	0.15
Male, n (%)	115 (68)	136 (69)	0.91
BSA, m2	1.63 [1.48–1.72]	1.63 [1.51–1.76]	0.35
INTERMACS profile, n (%)			<0.001
1–2	8 (5)	42 (21)
3	93 (55)	107 (55)
4–7	15 (9)	9 (5)
Bridge-to-bridge	52 (31)	38 (19)
Bridge device, n (%)
Extracorporeal VAD	33 (20)	34 (17)	0.59
Impella	19 (11)	4 (2)	0.0003
Intended use of device, n (%)			<0.001
Destination therapy	26 (15)	3 (2)
Bridge-to-transplant	137 (82)	194 (99)
Bridge-to-candidacy for transplant	5 (3)	0
Medical history, n (%)
Hypertension	28 (17)	23 (14)	0.23
Hyperlipidemia	59 (35)	64 (33)	0.66
Diabetes	32 (19)	53 (27)	0.08
Etiology of heart failure, n (%)			0.31
DCM	79 (47)	112 (57)
ICM	29 (17)	29 (15)
dHCM	23 (14)	25 (13)
Myocarditis	12 (7)	12 (6)
Other	25 (15)	18 (9)
Laboratory data
Hb, g/dL	11.6 [9.9–13.2]	11.6 [9.9–13]	0.57
Creatinine, mg/dL	0.89 [0.70–1.17]	0.93 [0.72–1.16]	0.64
eGFR, mL/min/1.73 m2	67.1 [48.4–85.4]	67.5 [48.0–91.7]	0.99
Alb, g/dL	3.7 [3.3–4.1]	3.8 [3.3–4.1]	0.64
PT-INR	1.3 [1.1–2.0]	1.2 [1.4–1.9]	0.28
BNP, pg/mL	405 [194–627]	377 [220–736]	0.48
Echocardiography
LVDd, mm	65 [56–76]	69 [61–77]	0.04
LVDs, mm	60 [51–72]	63 [55–72]	0.11
LVEF, %	21 [16–27]	18 [13–24]	<0.0001
TRPG, mmHg	26 [21–39]	30 [22–41]	0.26
Aortic regurgitation
Mild	56 (33)	44 (22)	0.03
≥Moderate	2 (1)	5 (3)	0.46
Tricuspid regurgitation
Mild	104 (62)	131 (67)	0.38
≥Moderate	16 (10)	21 (11)	0.73
Right heart catheterization
Cardiac index, L/min/m2	1.94 [1.71–2.4]	2.05 [1.72–2.49]	0.45
Mean PAP, mmHg	24 [16–30]	26 [17–35]	0.04
PCWP, mmHg	16 [9–22]	18 [10–26]	0.03
Mean RAP, mmHg	6 [3–8]	6 [4–10]	0.23
RVSWI	430 [274–600]	466 [330–646]	0.06
PAPi	3.2 [2.0–5.0]	3.0 [2.1–5.0]	0.88
PVR	2.3 [1.5–3.4]	2.3 [1.6–3.4]	0.60

Alb, albumin; BNP, B-type natriuretic peptide; BSA, body surface area; BTB, bridge-to-bridge; DCM, dilated cardiomyopathy; dHCM, dilated phase hypertrophic cardiomyopathy; DT, destination therapy; eGFR, estimated glomerular filtration rate; Hb, hemoglobin; ICM, ischemic cardiomyopathy; INTERMACS, Interagency Registry for Mechanically Assisted Circulatory Support; LVDd, left ventricular end-diastolic dimension; LVDs, left ventricular end-systolic dimension; LVEF, left ventricular ejection fraction; PAP, pulmonary artery pressure; PAPi, pulmonary artery pulsatility index; PCWP, pulmonary capillary wedge pressure; PT-INR, prothrombin time-international normalized ratio; PVR, pulmonary vascular resistance; RAP, right atrial pressure; RVSWI, right ventricular stroke work index; TRPG, tricuspid regurgitation pressure gradient; VAD, ventricular assist device.

Operative and Perioperative Outcomes

Operative and perioperative outcomes are presented in Table 2. The median operative time was significantly shorter in the HM3 group than in the HMII group (261 [208–347] vs. 330 [242–424] min, P<0.0001), whereas CPB time was similar between groups (P=0.89). Rates of reoperation (35% vs. 32%, P=0.50), tricuspid valve procedures (26% vs. 29%, P=0.56), and temporary RVAD implantation (7% vs. 4%, P=0.23) were also comparable. Aortic valve procedures were more frequently performed in the HM3 group (23% vs. 13%, P=0.01). The HM3 group required greater volumes of red blood cell (RBC) transfusion (16 vs. 12 units, P<0.0001), but platelet concentrate transfusion was significantly lower (20 vs. 35 units, P=0.0004). Fresh frozen plasma usage (P=0.10) and re-exploration for bleeding (17% vs. 22%, P=0.35) were not significantly different. Postoperative echocardiography showed smaller LVDd and left ventricular end-systolic dimension in the HM3 group. Tricuspid regurgitation pressure gradient values were comparable between groups (P=0.87). The prevalence of aortic and tricuspid regurgitation did not differ significantly between groups. The cardiac index was slightly lower in the HM3 group (2.5 vs. 2.6 L/min/m², P=0.03), but other hemodynamic parameters were similar. In-hospital mortality rates were not significantly different (7% vs. 5%, P=0.65); however, the hospital stay period was significantly shorter in the HM3 group (87 [69–125] vs. 98 [78–132] days, P=0.01).

Table 2.

Operative and Perioperative Outcomes

Variables	HeartMate 3 (n=168)	HeartMate II (n=196)	P value
Redo surgery, n (%)	59 (35)	62 (32)	0.50
Operation time, min	261 [208–347]	330 [242–424]	<0.0001
CPB time, min	106 [76–144]	104 [76–147]	0.89
Concomitant procedures
Aortic valve procedure, n (%)	39 (23)	26 (13)	0.01
Tricuspid valve procedure, n (%)	43 (26)	56 (29)	0.56
Temporary RVAD implantation, n (%)	11 (7)	7 (4)	0.23
Transfusion, Units
RBC	16 [10–22]	12 [6–18]	<0.0001
FFP	18 [12–24]	15 [10–24]	0.10
PC	20 [20–40]	35 [20–40]	0.0004
NO usage, n (%)	98 (58)	81 (44)	0.006
Re-exploration for bleeding, n (%)	29 (17)	42 (22)	0.35
Echocardiography
LVDd, mm	49 [41–60]	54 [46–62]	0.02
LVDs, mm	43 [34–54]	49 [40–57]	0.01
LVEF, %	23 [15–30]	15 [13–20]	<0.0001
TRPG, mmHg	17 [14–21]	17 [14–20]	0.87
Aortic regurgitation
Mild	59 (35)	64 (33)	0.66
≥Moderate	0	3 (2)	0.25
Tricuspid regurgitation
Mild	85 (51)	108 (55)	0.40
≥Moderate	4 (2)	6 (3)	0.76
Right heart catheterization
Cardiac index, L/min/m2	2.5 [2.3–2.8]	2.6 [2.3–3.1]	0.03
Mean PAP, mmHg	14 [12–17]	14 [12–17]	0.52
PCWP, mmHg	6 [4–9]	5 [3–8]	0.08
Mean RAP, mmHg	6 [4–8]	5 [3–7]	0.05
Hospital death, n (%)	11 (7)	10 (5)	0.65
Hospital stay period, days	87 [69–125]	98 [78–132]	0.01

CPB, cardiopulmonary bypass; FFP, fresh frozen plasma; NO, nitric oxide; PC, platelet concentrate; RBC, red blood cell; RVAD, right ventricular assist device. Other abbrevations as in Table 1.

Survival and Composite Endpoints

The 5-year event-free survival from the composite endpoint, defined as survival with LVAD support, freedom from stroke, and freedom from pump replacement, was significantly higher in the HM3 group than in the HMII group (75% vs. 52%; hazard ratio [HR] 0.52, 95% confidence interval [CI] 0.35–0.77; log-rank P=0.001) (Figure 1A).

Figure 1.

Kaplan-Meier analysis of composite endpoint and survival. (A) Freedom from the composite endpoint, defined as survival on LVAD support without stroke or pump replacement. (B) Overall survival following LVAD implantation. CI, confidence interval; HR, hazard ratio; LVAD, left ventricular assist device.

In contrast, overall survival at 5 years did not significantly differ between groups (90% in the HM3 group vs. 85% in the HMII group; HR 0.80, 95% CI 0.42–1.53; log-rank P=0.44) (Figure 1B). In the HM3 group, 15 patients died: 5 from infections, 4 from stroke, 5 from other organ failure, and 1 from airway bleeding. In the HMII group, 27 patients died: 21 from stroke, 3 from infections, 2 from heart failure, and 2 from unknown causes. Multivariable Cox regression analysis (Table 3) demonstrated that HM3 support was independently associated with a reduced risk of the 5-year composite outcome (HR 0.47, 95% CI 0.30–0.72, P=0.0006). However, HM3 support was not significantly associated with 5-year all-cause death (HR 0.75, 95% CI 0.39–1.44, P=0.39). In this model, redo surgery (HR 3.81, 95% CI 1.64–8.86, P=0.002) and lower serum albumin (HR 0.46, 95% CI 0.26–0.79, P=0.005) were identified as independent predictors of death, whereas extracorporeal VAD use was not significantly associated with death after adjustment.

Table 3.

Cox Proportional Hazards Regression Analysis for 5-Year Composite Endpoint and Mortality

Variable	Univariable			Multivariable
	HR	95% CI	P value	HR	95% CI	P value
(A) Cox proportional hazards regression analysis for 5-year composite endpoint
Age, per 10 years	1.11	0.96–1.28	0.16
Male	0.99	0.67–1.48	0.97
Diabetes mellitus	1.52	0.98–2.35	0.06
INTERMACS profile 1–2	0.86	0.50–1.48	0.59
Extracorporeal VAD	2.19	1.46–3.30	0.0002	0.68	0.35–1.32	0.26
Impella	0.70	0.28–1.70	0.43
DT	0.88	0.38–2.00	0.76
DCM	0.74	0.52–1.07	0.11
dHCM	0.79	0.44–1.41	0.42
ICM	1.88	1.22–2.90	0.004	1.64	1.02–2.64	0.04
Myocarditis	1.07	0.50–2.29	0.87
Hb, per 1 g/dL	0.83	0.75–0.91	0.0001	0.89	0.80–1.00	0.04
Alb, per 1 g/dL	1.01	0.99–1.01	0.32
eGFR, per 1 mL/min/1.73 m2	1.00	0.99–1.00	0.97
BNP, per 1 pg/mL	1.00	1.00–1.01	0.47
LVDd, per 1 mm	0.98	0.97–1.00	0.02	1.00	0.98–1.01	0.58
Redo surgery	2.40	1.67–3.46	<0.0001	2.39	1,45–3.92	0.0006
HM3 vs. HMII	0.52	0.35–0.77	0.001	0.47	0.30–0.72	0.0006
(B) Cox proportional hazards regression analysis for 5-year mortality
Age, per 10 years	1.27	0.98–1.67	0.08
Male	1.75	0.80–3.80	0.16
Diabetes mellitus	1.09	0.46–2.57	0.84
INTERMACS profile 1–2	1.16	0.48–2.77	0.74
Extracorporeal VAD	3.47	1.83–6.57	0.0001	0.77	0.32–1.89	0.57
Impella	0.41	0.06–3.02	0.38
DT	1.32	0.40–4.34	0.64
DCM	0.41	0.21–0.80	0.009
dHCM	1.46	0.64–3.30	0.36
ICM	2.31	1.15–4.63	0.02
Myocarditis	1.86	0.66–5.23	0.24
Hb, per 1 g/dL	0.73	0.60–0.86	0.0003
Alb, per 1 g/dL	0.36	0.22–0.60	<0.0001	0.46	0.26–0.79	0.005
eGFR, per 1 mL/min/1.73 m2	1.00	1.00–1.01	0.85
BNP, per 1 pg/mL	1.00	1.00–1.00	0.64
LVDd, per 1 mm	0.96	0.94–0.99	0.002
Redo surgery	3.84	2.00–7.40	<0.0001	3.81	1.64–8.86	0.002
HM3 vs. HMII	0.80	0.42–1.52	0.49	0.75	0.39–1.44	0.39

CI, confidence interval; HR, hazard ratio. Other abbrevations as in Table 1.

Each Adverse Event

The 5-year freedom from stroke was significantly higher in the HM3 group than in the HMII group (87% vs. 72%; HR 0.40, 95% CI 0.23–0.69; P=0.0008) (Figure 2A). During the study period, 17 stroke cases were observed in the HM3 group, including 12 hemorrhagic and 5 ischemic stroke, and of these, 6 were classified as severe, defined by a modified Rankin Scale score of ≥ 4. The 5-year freedom from bleeding was also significantly higher in the HM3 group (94% vs. 77%; HR 0.22, 95% CI 0.10–0.46; P<0.0001) (Figure 2B). Specifically, the 5-year freedom from gastrointestinal bleeding was 96% in the HM3 group and 89% in the HMII group (HR 0.65, 95% CI 0.53–0.81; P=0.042) (Figure 2C). The 5-year freedom from pump replacement due to any reason was significantly higher in the HM3 group (92% vs. 72%; HR 0.36, 95% CI 0.18–0.73; P=0.003) (Figure 2D). When categorized by etiology, freedom from pump replacement due to thrombosis was 99% in the HM3 group and 90% in the HMII group (HR 0.09, 95% CI 0.01–0.65; P=0.003) (Figure 2E). Freedom from pump replacement due to infection was comparable between groups (93% vs. 92%; HR 1.31, 95% CI 0.52–3.33; P=0.56) (Figure 2F). In contrast, freedom from pump replacement due to driveline fracture was significantly higher in the HM3 group (100% vs. 87%; P=0.003) (Figure 2G). The 5-year freedom from driveline infection was comparable between groups (63% in HM3 vs. 68% in HMII; HR 1.16, 95% CI 0.78–1.72; P=0.46) (Figure 2H). Late RHF occurred at similar rates in both groups, with a 5-year freedom of 91% in the HM3 group and 87% in the HMII group (HR 0.91, 95% CI 0.45–1.84; P=0.79) (Figure 2I). Freedom from unplanned rehospitalization for LVAD-related events was higher in the HM3 group (41% vs. 31%; HR 0.82, 95% CI 0.61–1.09; P=0.17) (Figure 2J). The most common cause of rehospitalization in the HM3 group was driveline infection, followed by heart failure, stroke, and arrhythmias. Regarding the occurrence of de-novo aortic regurgitation requiring surgical reoperation, freedom from reoperation was similar between the HMII and HM3 groups (Supplementary Figure 2).

Figure 2.

Kaplan-Meier analysis of freedom from adverse events. (A) Stroke, (B) bleeding, (C) gastrointestinal bleeding, (D) pump replacement due to any reasons, (E) pump replacement due to pump thrombosis, (F) pump replacement due to pump infection, (G) pump replacement due to driveline fracture, (H) driveline infection, (I) late right heart failure, and (J) rehospitalization for LVAD-related events. CI, confidence interval; HR, hazard ratio; LVAD, left ventricular assist device; RHF, right heart failure.

To further clarify the clinical impact of device type in the chronic phase, stroke and bleeding events were additionally analyzed in patients who maintained LVAD support beyond 1 year. In this subset analysis, freedom from stroke and bleeding after the first year of implantation remained higher in the HM3 group, although the differences were not statistically significant (stroke: HR 0.79, 95% CI 0.35–1.79, P=0.32; bleeding: HR 0.68, 95% CI 0.20–2.28, P=0.53) (Figure 3A,B).

Figure 3.

Kaplan-Meier analysis of (A) late stroke and (B) bleeding events in patients with continued LVAD support beyond 1 year.

Discussion

In this multicenter retrospective study, comparing long-term outcomes of HM3 and HMII support in Japanese patients revealed that: (1) although the overall 5-year survival was comparable between groups, the 5-year composite event-free survival was significantly higher in the HM3 group; (2) the incidence of major adverse events, including stroke and bleeding, was significantly lower in the HM3 group, especially during the acute phase; and (3) despite a lower incidence of some adverse events with the HM3, rates of unplanned rehospitalization remained comparable between the 2 devices.

The 5-year composite outcome, defined as survival with LVAD support without stroke or pump replacement, was significantly improved in patients receiving the HM3, which aligns with findings from previous studies, including the MOMENTUM 3 trial, and supports the clinical benefit of HM3 in reducing serious device-related complications.^8,9,12,17 Multivariable Cox regression analysis further confirmed that HM3 support was independently associated with improved composite outcomes. The long-term outcomes of HM3 support in our Japanese cohort appeared more favorable than those reported in MOMENTUM 3. These differences are likely attributable to baseline characteristics, including younger patient age, a lower proportion of DT indication, and prolonged transplant waiting times in Japan, which result in clinically stable BTT patients remaining on LVAD support for several years. Therefore, our findings should be interpreted as reflective of this unique clinical setting, rather than as indicating direct superiority over international cohorts.

One of the most notable advantages of the HM3 observed in this study was the significant reduction in device-related adverse events associated with hemocompatibility. Stroke and pump thrombosis occurred less frequently in the HM3 group, consistent with a previous report.¹⁹ In our cohort, pump thrombosis was rare, and the incidence of stroke was lower, with most events being hemorrhagic. In addition, bleeding events not involving the brain, such as gastrointestinal bleeding, were less common in the HM3 group. These findings suggest that the intrinsic hemocompatibility and improved flow dynamics of the HM3 help to minimize blood trauma and maintain hemodynamic stability.^11,15 Notably, HM3 support was associated with a significantly lower incidence of thromboembolic and bleeding events, particularly during the early postoperative phase. These adverse events occur more frequently within the first year after implantation,²⁰ potentially due to factors such as intraoperative depletion of coagulation factors and unstable anticoagulation control with warfarin and heparin during the early postoperative period. Despite significantly higher intraoperative transfusion requirements, especially for RBC, in the HM3 group, the incidence of thrombotic and hemorrhagic complications remained low. This increased transfusion demand may be attributed to the older patient population, a higher rate of Impella bridging, and more frequent concomitant aortic valve procedures. Nevertheless, the low rate of adverse events under these challenging conditions is consistent with the favorable hemocompatibility and safety profile of the HM3. Notably, operative times were shorter in the HM3 group despite a higher frequency of concomitant aortic valve procedures. This difference is likely attributable to substantial surgical experience and to implantation know-how accumulated during the HMII era. In addition, HM3 implantation does not require the extensive pocket dissection necessary for HMII, which may have also contributed to shorter operative times.

Though some adverse events were reduced with HM3, infection rates, including driveline and pump infections, and the incidence of late RHF were similar between groups. Unplanned rehospitalization rates also remained comparable between groups. The absence of a significant difference in LVAD-related infection rates has been reported in previous studies.^12,21 One possible explanation for the similarity in rehospitalization rates is that a considerable number of rehospitalizations in both groups were due to driveline infections. This observation underscores the persistent clinical challenge of infection prevention. Enhanced strategies for patient education, surgical technique, and postoperative care at the driveline exit site are needed. Similarly, the incidence of late RHF was comparable between groups. Despite technological improvements in the HM3 device, its support did not reduce the risk of this complication,¹² which indicates that RHF remains a complex clinical challenge and emphasizes the importance of careful perioperative management and individualized proper ventricular assessment. Nevertheless, the reduced incidence of reoperations and strokes associated with HM3 support may still translate into benefits for quality of life, even though overall rehospitalization rates were unchanged.

Study Limitations

First, it was a retrospective observational study without a prespecified analysis plan, therefore, the statistical analyses should be regarded as exploratory. Second, the implantation periods for HMII and HM3 were markedly different; era-related improvements in patient selection, perioperative management, and institutional experience, as well as differences in the use of temporary mechanical circulatory support such as Impella, may have influenced outcomes independent of device characteristics. Third, during the HMII era, multiple implantable LVADs were available and were used in some cases, whereas during the HM3 era, the HM3 was the sole device, which may have affected patient selection. Fourth, although we considered propensity score matching, the relatively small sample size and substantial era-related differences in management precluded a reliable analysis. Finally, detailed patient-level data regarding the dose, duration, and longitudinal use of guideline-directed medical therapy were not consistently collected in this retrospective dataset, limiting our ability to assess the impact of post-implantation pharmacotherapy on long-term outcomes.

Conclusions

HM3 support was associated with a significantly lower incidence of device-related complications, including stroke, bleeding, pump thrombosis, and driveline failure, compared to HMII. Although overall 5-year survival was comparable between groups, HM3 support provided improved freedom from the composite endpoint of survival without major device-related complications. These findings support the clinical advantage of the HM3 in long-term mechanical circulatory support and suggest that, in Japan where prolonged transplant waiting times necessitate long-term BTT, the HM3 may contribute to safer and more sustainable LVAD management.

Acknowledgments

During the preparation of this manuscript, the authors used ChatGPT (OpenAI) to improve English readability and clarify phrasing. We thank Editage for their writing support while drafting this manuscript.

Disclosures

Y.S. is a member of Circulation Journal’s Editorial Team.

Conflict of Interest

The authors declare they have no conflicts of interest.

Funding Sources

This research received no specific grants from any funding agency in the public, commercial, or not-for-profit sectors.

IRB Information

This study was approved by the institutional review board of Osaka University Hospital (approval number: 21372(T1), approval date: January 18, 2022).

Author Contributions

All authors contributed to the conception and design of the study. T.T., Y.Y., K.T., and N. Kitahata contributed to the data collection. T.T. and D.Y. performed data analysis. T.T. and D.Y. drafted the manuscript, which all the authors critically revised. All authors read and approved the final version of the manuscript for submission.

Data Availability

The data that support the findings of this study are available from the author, T.T., upon reasonable request.

Consent for Publication

This study was approved by the institutional review board. All the patients gave written informed consent for the use of their medical records.

Supplementary Files

Please find supplementary file(s);

https://doi.org/10.1253/circj.CJ-25-0597

References

• 1.   Ran J, Zhou P, Wang J, Zhao X, Huang Y, Zhou Q, et al. Global, regional, and national burden of heart failure and its underlying causes, 1990–2021: Results from the global burden of disease study 2021. Biomark Res 2025; 13: 16, doi:10.1186/s40364-025-00728-8.
• 2.   Crespo-Leiro MG, Metra M, Lund LH, Milicic D, Costanzo MR, Filippatos G, et al. Advanced heart failure: A position statement of the Heart Failure Association of the European Society of Cardiology. Eur J Heart Fail 2018; 20: 1505–1535, doi:10.1002/ejhf.1236.
• 3.   Roesel MJ, Nersesian G, Neuber S, Thau H, Wolff von Gudenberg R, Lanmueller P, et al. LVAD as a bridge to transplantation: Current status and future perspectives. Rev Cardiovasc Med 2024; 25: 176, doi:10.31083/j.rcm2505176.
• 4.   Melendo-Viu M, Dobarro D, Raposeiras Roubin S, Llamas Pernas C, Moliz Cordon C, Vazquez Lamas M, et al. Left ventricular assist device as a destination therapy: Current situation and the importance of patient selection. Life (Basel) 2023; 13: 1065, doi:10.3390/life13041065.
• 5.   Kitai T, Kohsaka S, Kato T, Kato E, Sato K, Teramoto K, et al. JCS/JHFS 2025 Guideline on diagnosis and treatment of heart failure. Circ J 2025; 89: 1278–1444, doi:10.1253/circj.CJ-25-0002.
• 6.   Seguchi O, Kuroda K, Kumai Y, Nakajima S, Yanase M, Wada K, et al. Clinical outcomes of patients with the HeartMate II left ventricular assist device: A single-center experience from Japan. Transplant Proc 2018; 50: 2726–2732, doi:10.1016/j.transproceed.2018.03.091.
• 7.   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, doi:10.1253/circj.CJ-17-0881.
• 8.   Mehra MR. The burden of haemocompatibility with left ventricular assist systems: A complex weave. Eur Heart J 2019; 40: 673–677, doi:10.1093/eurheartj/ehx036.
• 9.   Colombo PC, Mehra MR, Goldstein DJ, Estep JD, Salerno C, Jorde UP, et al. Comprehensive analysis of stroke in the long-term cohort of the MOMENTUM 3 study. Circulation 2019; 139: 155–168, doi:10.1161/CIRCULATIONAHA.118.037231.
• 10.   Bansal A, Uriel N, Colombo PC, Narisetty K, Long JW, Bhimaraj A, et al. Effects of a fully magnetically levitated centrifugal-flow or axial-flow left ventricular assist device on von Willebrand factor: A prospective multicenter clinical trial. J Heart Lung Transplant 2019; 38: 806–816, doi:10.1016/j.healun.2019.05.006.
• 11.   Netuka I, Kvasnicka T, Kvasnicka J, Hrachovinova I, Ivak P, Marecek F, et al. Evaluation of von Willebrand factor with a fully magnetically levitated centrifugal continuous-flow left ventricular assist device in advanced heart failure. J Heart Lung Transplant 2016; 35: 860–867, doi:10.1016/j.healun.2016.05.019.
• 12.   Mehra MR, Goldstein DJ, Cleveland JC, Cowger JA, Hall S, Salerno CT, et al. Five-year outcomes in patients with fully magnetically levitated vs axial-flow left ventricular assist devices in the MOMENTUM 3 randomized trial. JAMA 2022; 328: 1233–1242, doi:10.1001/jama.2022.16197.
• 13.   Mehra MR, Uriel N, Naka Y, Cleveland JC Jr, Yuzefpolskaya M, Salerno CT, et al. A fully magnetically levitated left ventricular assist device: Final report. N Engl J Med 2019; 380: 1618–1627, doi:10.1056/NEJMoa1900486.
• 14.   Uriel N, Colombo PC, Cleveland JC, Long JW, Salerno C, Goldstein DJ, et al. Hemocompatibility-related outcomes in the MOMENTUM 3 trial at 6 months: A randomized controlled study of a fully magnetically levitated pump in advanced heart failure. Circulation 2017; 135: 2003–2012, doi:10.1161/CIRCULATIONAHA.117.028303.
• 15.   Vandenberge J, Han D, Sun W, Wang S, Tran D, Kim N, et al. In vitro comparison of device-induced hemolysis, platelet defects, and von Willebrand factor degradation between the HeartMate 2 and HeartMate 3 pumps. Artif Organs 2025; 49: 1380–1386, doi:10.1111/aor.15013.
• 16.   Fang P, Du J, Boraschi A, Bozzi S, Redaelli A, Schmid Daners M, et al. Insights into the low rate of in-pump thrombosis with the HeartMate 3: Does the artificial pulse improve washout? Front Cardiovasc Med 2022; 9: 775780, doi:10.3389/fcvm.2022.775780.
• 17.   van der Heiden CW, Zijderhand CF, Veen KM, Constantinescu AA, Manintveld OC, Brugts JJ, et al. Mechanical device malfunction of the HeartMate II versus the HeartMate 3 left ventricular assist device: The Rotterdam experience. ASAIO J 2023; 69: e80–e85, doi:10.1097/MAT.0000000000001877.
• 18.   Jorde UP, Saeed O, Koehl D, Morris AA, Wood KL, Meyer DM, et al. The Society of Thoracic Surgeons Intermacs 2023 annual report: Focus on magnetically levitated devices. Ann Thorac Surg 2024; 117: 33–44, doi:10.1016/j.athoracsur.2023.11.004.
• 19.   Inoue K, Fujita T, Yoshioka D, Tonai K, Yanagino Y, Kakuta T, et al. Short-term outcomes of magnetically levitated left ventricular assist device in advanced heart failure: The Japanese cohort. Circ J 2022; 86: 1961–1967, doi:10.1253/circj.CJ-22-0332.
• 20.   Patel CB, Blue L, Cagliostro B, Bailey SH, Entwistle JW, John R, et al. Left ventricular assist systems and infection-related outcomes: A comprehensive analysis of the MOMENTUM 3 trial. J Heart Lung Transplant 2020; 39: 774–781, doi:10.1016/j.healun.2020.03.002.
• 21.   Kirklin JK, Pagani FD, Kormos RL, Stevenson LW, Blume ED, Myers SL, et al. Eighth annual INTERMACS report: Special focus on framing the impact of adverse events. J Heart Lung Transplant 2017; 36: 1080–1086, doi:10.1016/j.healun.2017.07.005.