For patients with end-stage heart failure, heart transplantation (HTx) is the gold standard of therapy. However, patients are increasingly bridged to HTx with a circulatory support system [1]. Most commonly, a durable left ventricular assist device (d-LVAD), which needs to be implanted surgically, has been used for bridging [1]. It unloads the left ventricle through the pump inflow tract implanted through the apex, pumping the blood back into the aorta through the anastomosed outflow tract. Meanwhile, temporary left ventricular assist devices (t-LVAD), which can be inserted percutaneously through the femoral or axillary artery and provide adequate flow rates to maintain circulation in the body, are available in the market. T-LVADs are inserted into the left ventricle through the aorta, and a microaxial pump unloads the left ventricle. These t-LVADs, even when approved only for temporary support, are also increasingly used for bridging the patient to HTx [2]. Finally, veno-arterial extracorporeal membrane oxygenation (VA-ECMO), consisting of an extracorporeal circuit with a pump and an oxygenator, is also used for bridging patients to HTx [1].

Besides bridging to HTx, primary graft dysfunction (PGD) after HTx is another indication of the implantation of a circulatory support system [3, 4]. Currently, a high level of evidence regarding all three circulatory support device options, d-LVAD, t-LVAD, and VA-ECMO, for bridging to HTx and treating PGD does not exist. Thus, this meta-analysis aimed to investigate the survival of patients after HTx supported with d-LVAD, t-LVAD, or VA-ECMO either for bridging to HTx or for treatment of PGD.

We included four studies for the bridging of HTx analysis and three studies for the PGD after HTx analysis. Reasons for exclusion are shown in the PRISMA flow chart in Fig. 1. The main reasons for exclusion were based on the patient population, device type, study design, and publication type in both analyses. In the selection process of studies for bridging to HTx, we also excluded studies that were based on data from overlapping time frames from the same registry.

Fig. 1.

PRISMA flow chart. (A) Bridging to HTx. (B) PGD after HTx. HTx, Heart transplantation; PGD, Primary graft dysfunction; PRISMA, Preferred Reporting Items for Systematic Reviews and Meta-analysis statement.

For both meta-analyses, we collected data regarding device-, recipient-, donor-, matching-, and transplant-related characteristics (Table 1, Ref. [2, 11, 12, 13] and Table 2, Ref. [14, 15, 16]). However, not every publication contained information on every data item we searched for. However, besides the data items shown in Tables 1,2, we also searched for multiple other data items that were unfortunately not shown in any of the publications. For the bridging to transplantation analysis, this included the recipient characteristics, total bilirubin, cytomegalovirus (CMV) serology pretransplant infection, blood transfusion while on waitlist, and cause of death if death occurred after transplantation. Unstated donor characteristics were donor type (brain death or circulatory), Graft LVEF 50%, total bilirubin, serum creatinine, graft maintenance method (static cold storage and normothermic blood perfusion), graft out-of-body time, and graft blood perfusion time in case of NRP. For the PGD after transplantation analysis, additional unstated items were race, blood type, diabetes mellitus, pretransplant mechanical ventilation, and pretransplant infection regarding recipient characteristics. Regarding donor characteristics, additional missing data items were race, diabetes mellitus, CMV serology, blood type, graft cold ischemic time, and graft maintenance method. Furthermore, the following matching-related characteristics were not given in studies regarding PGD: sex-, race-, HLA-, AB0-, and CMV-matched.

3.1 Bridging to Heart Transplantation

3.1.1 Device-Related Characteristics

We included two studies, which reported results after using a d-LVAD, one VA-ECMO study, and one t-LVAD study (Table 1). The t-LVAD study reported characteristics and results separately using the Impella 5.0 and the Impella 5.5 device [2]. Thus, we included each subgroup as a separate study in the analysis. Kilcoyne et al.’s registry study [11] included only patients supported with an HM3 LVAD. In contrast, the single-center study by Bedanova et al. [13] included patients supported with an HM2 or HW. In total, n = 1678 patients were included in the analysis.

3.1.2 Recipient Characteristics

The d-LVAD study by Kilcoyne et al. [11] provided only information on device-related characteristics. The mean recipient age ranged between 44.9 years in the VA-ECMO study and 53 to 58 years in the d-LVAD and t-LVAD studies (Table 1). Both t-LVAD populations were predominantly male (80.9% and 86.9%). In the d-LVAD study by Bedanova et al. [13], only 45% of patients were male. The body mass index was comparable between the studies and ranged from 25.5 to 27.7 kg/m². Information regarding the blood type, pretransplant mechanical ventilation, and intravenous inotropes at registration cannot be compared between studies due to a lack of reporting. Diabetes mellitus was prevalent, with 11.3% in the VA-ECMO study, 17% in the d-LVAD study by Bedanova et al. [13], and 31.1 to 33.8% in the t-LVAD studies.

3.1.3 Donor, Matching- and Transplant-Related Characteristics

The donor age was 31 to 35 years in the t-LVAD and VA-ECMO study and 42 years in the d-LVAD study by Bedanova et al. [13]. The prevalence of diabetes mellitus in the donor was low, as stated, 2.4 to 5.2% in the VA-ECMO and t-LVAD studies. Neither of the d-LVAD studies states any matching characteristics. Compared to the t-LVAD studies, sex-matching was low in the VA-ECMO study, with 62.7%. The graft cold ischemic time was 202 to 208 min in the t-LVAD studies and 222 min in the VA-ECMO study.

3.1.4 Outcome

After HTx in patients bridged to transplantation with a circulatory support system, the outcome was presented homogeneously as 1-year overall survival in all included studies (Fig. 2). The 1-year survival was 94.6% in the Impella 5.5 group and 91.3% in the Impella 5.0 group, leading to a survival of 92.7% (95% CI: 89.2 to 95.6%) after HTx in patients bridged with a t-LVAD. The 1-year survival was 80.0% and 90.4% in the d-LVAD studies, leading to a 1-year survival after HTx in patients bridged with a d-LVAD of 86.8% (95% CI: 75.8 to 94.8%). Survival after HTx in patients bridged with a VA-ECMO was 71.6% (95% CI: 63.7 to 78.9%). The 1-year survival in all included studies was 86.9% (95% CI: 78.2 to 93.7%). The statistical comparison between d-LVAD, t-LVAD, and VA-ECMO revealed that bridging with a d-LVAD (p = 0.027) and t-LVAD (p = 0.001) seems to result in a significantly higher 1-year survival probability after HTx than bridging with VA-ECMO. The 1-year survival probability of d-LVAD and t-LVAD was not significantly different (p = 0.210).

Fig. 2.

One-year survival after heart transplantation in patients bridged with a circulatory support system. HM2, HeartMate 2; HM3, HeartMate 3; HW, HeartWare; VA-ECMO, Venoarterial extracorporeal membrane oxygenation; LVAD, Left ventricular assist device; d-LVAD, durable LVAD; t-LVAD, temporary LVAD; LVEF, Left ventricular ejection fraction; HW, HeartWare; RE, Random effects; CI, Confidence interval.

3.2 Primary Graft Dysfunction After Heart Transplantation

3.2.1 Device-Related Characteristics

For treating PGD after HTx, we included three studies in total (Table 2). Two studies reported results after circulatory support with VA-ECMO, and one study reported results after circulatory support of two cases with a t-LVAD, namely CentriMag. In total, n = 35 patients were included.

3.2.2 Recipient Characteristics

The t-LVAD study did not report information regarding donor and recipient characteristics. The recipient age ranged from 47.6 to 55.1 years in the VA-ECMO studies. Male sex ranged from 72 to 100%. The body mass index was 22.8 kg/m² in the study by Yuan et al. [16] and 31.7 kg/m² in the study by Kawabori et al. [14]. The most recent postoperative creatinine was 1.46 and 1.76 g/dL.

3.2.3 Outcome

The outcome after HTx in patients with PGD treated with a circulatory support system was presented homogeneously as 30-day survival in all included studies (Fig. 3). Based on the very limited data, the 30-day survival rate in the study by Yuan et al. [16] was 72.5%, equivalent to 17.4 of 24 patients. Based on the survival curve, which was also provided in the publication, we assumed a 75% 30-day survival, which would correspond to 18 of 24 patients. The 30-day survival rate in patients who were treated with t-LVAD was 100% (95% CI: 59.2 to 100%). In patients supported with VA-ECMO for PGD, the outcome was 75% and 100%, leading to a 30-day survival of 92.4% (95% CI: 66 to 100%).

Fig. 3.

30-day survival in patients with primary graft failure after heart transplantation treated with a circulatory support system. VA-ECMO, Venoarterial extracorporeal membrane oxygenation; t-LVAD, temporary left ventricular assist device; RE, Random effects; CI, Confidence interval.

For bridging to HTx, t-LVAD support seems to lead to at least equivalent 1-year survival rates after HTx as d-LVAD support. This might be because the implanted t-LVAD devices provide sufficient circulatory support comparable to a d-LVAD. Thus, both support types seem to stabilize the transplant candidate equally during bridging. However, the survival rate in patients bridged with VA-ECMO appears to be lower, although they were almost ten years younger than those in the other studies. A worse outcome in patients bridged with VA-ECMO compared to d-LVAD bridging is consistent with a published article by Lund et al. [4]. Nevertheless, a comparison with a t-LVAD has been missing so far because, at the time of the publication of the mentioned article, bridging with a t-LVAD had not been performed. This gap in knowledge has now been closed. ECMO therapy increases the risk of neurologic complications, bleeding, thrombotic events, renal failure, and vascular-access complications [17, 18, 19, 20, 21]. As a result, the status of transplant candidates bridged with VA-ECMO might be worse due to possibly decreased kidney and liver function or coagulation problems compared to patients who were bridged with a d-LVAD. This could possibly also affect the post-transplant outcome. Besides possible outcome differences between the circulatory support systems, the survival after HTx in general appears relatively high. A high survival after HTx can be based on an intended analysis of patients with a good outcome, young or carefully selected recipients, or improvements in donor or recipient management. This first aspect is unlikely to have appeared because the registry studies, which contributed to the high number of patients, described a well-elaborated inclusion protocol. The mean recipient age, when reported, was also not surprisingly low.

Nevertheless, the recipient age was unknown in the d-LVAD study by Kilcoyne et al. [11], which included over a thousand patients. Donor and recipient management are important for survival. However, the time frame of the included studies is very comparable, so no device-type-specific bias should have occurred.

The d-LVAD study by Bedanova et al. [13] showed a worse 1-year survival compared to the registry d-LVAD study and compared to the t-LVAD studies. However, it also needs to be mentioned that in the study by Bedanova et al. [13], the donors were about ten years older than the recipients bridged with a t-LVAD. Considering that donor age is a risk factor for 1-year post-transplant mortality, the higher donor age could explain the differing outcomes of the studies [4]. The t-LVAD study predominantly consists of male recipients, who might be less likely to develop graft failure than female recipients based on the literature [22, 23]. However, broad evidence that female sex is a risk factor for impaired early survival is missing. The impact of recipient diabetes mellitus, which was more prevalent in the t-LVAD study, on post-transplant survival has been investigated multiple times. However, based on the literature, conclusions remain unclear [24, 25]. It is also important to highlight that the d-LVAD study by Bedanova et al. [13] is single-center, while the t-LVAD studies are registry studies. Registry studies allow the inclusion of a large amount of data collected from multiple centers. However, the data quality depends on the individual center and thus might be heterogeneous [26]. In the case of a single-center study, data quality tends to be homogenous. At the same time, the therapy quality is based on the individual experience of a single center, and most likely, patient selection is restricted.

It would be interesting to reanalyze outcomes after HTx in patients bridged with a t-LVAD, which is intended and approved for longer use and thus more appropriate as a bridging device when available on the market. Finally, the impact of the support device on the high-urgency listing status of a transplant candidate might also have an impact, as only d-LVAD implantation leads to losing the high-urgency status.

Only a few studies exist on treating post-transplant PGD with a circulatory support system. Additionally, the number of patients is low, consisting of only two cases with a t-LVAD. Consequently, the interpretability of results regarding this part of the meta-analysis is reduced. Furthermore, data representing the outcome of patients after PGD treated with an Impella 5.0 or 5.5 system, which is one of the most frequently used t-LVADs nowadays, has not been published yet. Furthermore, two of three studies had almost no donor- and transplant-related information. Thus, donor- or recipient-associated factors that might interfere with the circulatory support system cannot be identified. The most recent postoperative creatinine was higher in the VA-ECMO study by Yuan et al. [16], which could partially explain the worse outcome [4]. The majority of included studies in both analyses are of a single-arm observational nature and can underlie a certain level of bias. Some of the included studies explicitly excluded multiorgan transplants, which leads to a stabilization of results. Nevertheless, in future studies, when larger data sets are available, inclusion of studies should be restricted to those that excluded multiorgan transplantation to exclude any possibility of bias.

D-LVAD, t-LVAD, and VA-ECMO are used to bridge patients to HTx. VA-ECMO and t-LVADs were used in the included studies to treat PGD post-HTx. The type of circulatory support device impacts the 1-year survival after HTx. Both d-LVAD and t-LVAD seem to result in a higher 1-year survival rate after HTx than VA-ECMO. However, d-LVAD and t-LVAD appear to have a comparable 1-year survival probability. Circulatory support for bridging to HTx is much more investigated, especially by registry studies, than for PGD after HTx. In PGD, VA-ECMO seems to be predominantly used. Unfortunately, the device-specific outcomes after the treatment of PGD cannot be compared due to low patient numbers. However, more prospective clinical studies investigating circulatory support devices in PGD after HTx need to be performed.

HTx, heart transplantation; d-LVAD, durable left ventricular assist device; t-LVAD, temporary left ventricular assist devices; VA-ECMO, veno-arterial extracorporeal membrane oxygenation; PGD, primary graft dysfunction; PRISMA, Preferred Reporting Items for Systematic Reviews and Meta-analysis statement.

Data is already included in this publication.

LS: Research design, investigation, data collection, analysis, interpretation, writing. SM: investigation, data collection, analysis. MK: investigation, analysis. AS: interpretation and review. GS: interpretation and review. All authors contributed to the critical revision of the manuscript for important intellectual content. All authors read and approved the final manuscript. All authors have participated sufficiently in the work and agreed to be accountable for all aspects of the work.

Not applicable.

Not applicable.

This project was supported by the German Research Foundation (Deutsche Forschungsgemeinschaft, DFG) to Dr. Lars Saemann (Project number: 530557324) and by EFRE (Project Thera4Age to Prof. Andreas Simm).

The authors declare no conflict of interest.

During the preparation of this work, the authors used Rayyan in order to organize electronic database results and automatically detect duplicates. After using this tool/service, the authors reviewed and edited the content as needed and take full responsibility for the content of the publication.

Supplementary material associated with this article can be found, in the online version, at https://doi.org/10.31083/RCM45064.