More than 250,000 patients in the United States suffer from end-stage heart failure refractory to maximized medical therapy.1 Heart transplantation has been established to improve quality and length of life in this population2; however, the efficacy of this treatment modality is limited by allograft availability.3,4 Because of the scarcity of donor organs, the use of mechanical circulatory support (MCS) as a bridge to transplant has been well established.5,6 The proportion of patients bridged to heart transplantation with MCS has steadily increased over the past 5 years, with more than one third of patients currently utilizing a mechanical support device at the time of transplant.7 Multiple options exist for the patient requiring MCS as a bridge to transplantation.6 Left ventricular assist devices (LVADs) are now the most commonly used device, currently representing approximately 89% of patients on support at the time of transplant.7
The dramatic increase in the proportion of heart transplant recipients bridged with LVAD therapy raises the possibility of a future where the majority of heart transplants will be preceded by placement of a ventricular assist device before transplantation.8 However, this approach results in obligate repeat sternotomy9 and may also increase bleeding risk because of anticoagulation and acquired von Willebrand disease10 as well as increased allosensitization.11,12 Whether these factors have a negative impact on posttransplant outcomes compared with less-invasive circulatory support strategies is unknown. Our objective was to examine the national experience with heart transplantation in the setting of less-invasive intraaortic balloon pump counterpulsation (IABP) versus intracorporeal left ventricular assist devices (dLVAD) at the time of surgery to determine whether dLVAD has a detrimental effect on posttransplant outcomes when compared with IABP. This must be considered in the context of pretransplant mortality and locoregional organ availability, which are also addressed.
This study was approved by the Institutional Review Board of Duke University Medical Center.
The United Network for Organ Sharing (UNOS) database encompassing all US heart transplant procedures performed between June 30, 2004, and December 31, 2011, with follow-up through March 31, 2012, was used for this study.
Adult (age, ≥18 years) heart transplant recipients were included for analysis. Multiorgan and repeat transplants were excluded. The study population was further limited to those treated with IABP or dLVAD at the time of transplantation (transplant recipients without circulatory support were included only for the purposes of calculating the proportion of IABP and dLVAD utilization). Cases where information regarding IABP or dLVAD use was missing or incomplete were also excluded.
Primary predictor variable.
The primary predictor variable for our analysis was the circulatory support modality at the time of transplant (IABP vs. dLVAD). Brands associated with durable, intracorporeal LVADs were included (Table 1).
Baseline characteristics and risk factors.
The UNOS database includes donor, recipient, and transplant-related characteristics. Please refer to Table 2 to for a listing of characteristics and risk factors extracted from the dataset for analysis.
Outcome measures and follow-up.
The primary outcome variable was overall posttransplant survival. Secondary outcome measures included perioperative mortality, postoperative length of stay >25 days, dialysis requirement before discharge, and acute rejection within the first postoperative year. Outcome data for each patient were ascertained from the date of transplantation until patient death, date of last follow-up, or the end of study period (March 31, 2012).
Trends in the use of IABP versus dLVAD over time as well as geographic usage patterns were reported descriptively. Baseline characteristics were described for the full study population using medians and 25th–75th percentile (Q1–Q3) for continuous variables and proportions (frequency, percentage) for discrete variables. Comparisons for continuous and ordered categorical variables were made using Kruskal–Wallis tests, and unordered categorical variables were compared using the Pearson χ2 test. Unadjusted patient survival rates were estimated using the product-limit (Kaplan–Meier) method and compared between groups by the log-rank test.
Multivariable Cox proportional hazards modeling was used to assess the simultaneous effect of circulatory support modality on risk of patient death, while adjusting for potential confounders based on covariates per the Scientific Registry of Transplant Recipients (SRTR) adult 3 year heart transplant risk model.13 These covariates are listed in the Supplemental Digital Content (http://links.lww.com/ASAIO/A90).
Univariable and multivariable binary logistic regression modeling was used to assess the simultaneous effect of circulatory support modality on risk of perioperative mortality, postoperative length of stay >25 days, dialysis requirement before discharge, and acute rejection within the first postoperative year (secondary endpoints), with risk adjustment again based on covariates as described earlier. Odds ratio and 95% confidence interval (CI) were calculated as measures of strength of association and precision, respectively.
Statistical analyses were performed using JMP version 10.0 (SAS Institute Inc., Cary, NC) and R version 2.15.1 (R Core Team 2012). For all comparisons, p values ≤0.05 were considered statistically significant. All p values are two sided.
Secondary Analysis—Pretransplant Mortality
In evaluating postoperative outcomes of IABP versus dLVAD at the time of transplant, it is important to put this information in context of pretransplant mortality. Therefore, a comparison of pretransplant mortality in patients with dLVAD versus IABP at the time of listing was performed. Please refer to Supplemental Digital Content (http://links.lww.com/ASAIO/A90).
Primary Study Population
A total of 3,207 patients meeting inclusion criteria having an IABP (n = 571, 17.8%) or dLVAD (n = 2,636, 82.2%) at transplant were included (see Figure 1 for study inclusion algorithm). The proportion of patients with a dLVAD at the time of transplant remained fairly constant between 2004 and 2008 at 12–15%, with more than doubling after 2008 to a peak of 31.1% in 2011 (Figure 2). The proportion of patients with an IABP at the time of transplant remained constant at 3.4–6.5% throughout the study period (Figure 2).
Geographic variation in IABP at transplant was observed (Figure 3). The highest proportion of patients supported with an IABP at the time of heart transplant was observed in the southeastern and northeastern states as well as Texas and Oklahoma (approximately 7%), whereas the lowest proportion of IABP use observed in the northwestern states as well as Michigan, Ohio, and Indiana (2–3%; Figure 3A). Conversely, the highest proportion of dLVAD use was observed in the northwestern states (30–35%) with the lowest proportion of dLVAD at the time of transplant in the southwestern states (approximately 10%; Figure 3B).
Baseline Recipient Characteristics
Compared with the dLVAD group, the IABP group had a higher proportion of females (20.7 vs. 16.7%, p = 0.023), higher creatinine at the time of transplant (median [Q1–Q3], 1.2 mg/dl [1.0–1.5] vs. 1.2 mg/dl [0.9–1.4], p = 0.001), lower body mass index (BMI; median [Q1–Q3], 25.8 [23.2–28.7] vs. 27.7 [24.4–31.1], p < 0.001), lower proportion with blood type O (35.4 vs. 49.7%, p < 0.001), and more functional impairment represented by a higher proportion requiring full assistance with activities of daily living at the time of listing for transplant (74.6 vs. 56.1%, p < 0.001; Table 2).
Baseline Donor and Transplant Characteristics
Baseline donor and transplant characteristics are summarized in Table 2. Donors for the IABP group demonstrated a lower BMI, higher proportion with cigarette use (>20 pack years ever), higher proportion with cocaine use, and a lower proportion requiring vasopressin administration within 24 hours of procurement (p ≤ 0.04 for all). Intraaortic balloon pump counterpulsation–supported recipients were predictably more likely to require intensive care, ventilator support, or inotropes at the time of transplant (p < 0.001 for all). Intraaortic balloon pump counterpulsation–supported recipients also experienced significantly shorter waitlist times (median [Q1–Q3], 22 [7–77] vs. 155 [63–314], p < 0.001).
The IABP group demonstrated significantly shorter unadjusted overall posttransplant survival, with separation in survival curves compared with dLVAD occurring at approximately 1.5 years posttransplant (Figure 4). On risk-adjusted multivariable Cox proportional hazard modeling, this relationship did not retain statistical significance (adjusted hazard ratio for IABP vs. dLVAD, 1.08; 95% CI, 0.87–1.33, p = 0.51; Figure 4).
On unadjusted comparison, the IABP group demonstrated a higher proportion with postoperative length of stay >25 days (23.9 vs. 19.8%, p = 0.033) and lower proportion with early acute rejection (20.9 vs. 25.5%, p = 0.048); however, on multivariable risk adjustment, there were no significant differences in secondary endpoints (p ≥ 0.14 for all; Table 3).
Secondary Analysis—Pretransplant Mortality with IABP Versus dLVAD at the Time of Listing
Please refer to Supplemental Digital Content, http://links.lww.com/ASAIO/A90, for results of the secondary analysis.
The dramatic increase in surgically implanted dLVAD use as a bridge to heart transplant underlines the importance of understanding how dLVAD therapy impacts posttransplant outcomes compared with less-invasive circulatory support modalities. Our study is one of the few contemporary analyses examining the use of IABP support as a bridge to heart transplant. By using national registry data, our results suggest that the presence of dLVAD at the time of heart transplantation does not negatively affect postoperative outcomes compared with IABP. This must be considered in the context of other factors impacting bridge-to-transplant management decisions, such as pretransplant mortality. These are important findings given the lower procedural risks associated with IABP support, which must be balanced against effectiveness of circulatory support and device-related complications.
Our data show that the strategy of IABP support is currently used for patients who have a smaller body habitus, a lower degree of allosensitization, or a non-O blood type. A greater proportion of transplant recipients who were supported with IABP were inotrope dependent, had ischemic disease as the etiology of heart failure, and had lower cardiac index. These recipient characteristics not only suggest more critically ill recipients but can also result in shorter waitlist times. Conversely, more patients in the dLVAD group were noted to be blood type O, and a greater proportion were allosensitized (primarily with class 1 anti-human leukocyte antigen antibodies), which would potentially require a long-term support device while awaiting transplantation. Given these differences, we found that the median wait time for patients supported with IABP was only 22 days compared with 155 days in the dLVAD group. However, this shorter wait time did not translate into improved posttransplant outcomes in the IABP group, which may be because of both donor and recipient factors. Although the choice between percutaneous versus durable MCS may at times be dictated by circumstances such as anatomic feasibility and expected wait times, there are many clinical circumstances where either modality would be a clinically reasonable approach as well as technically feasible, which in part relates to evolving technology.14 Therefore, the data provided in this study help guide clinical decisions in this evolving field.
The IABP group demonstrated significantly shorter unadjusted overall posttransplant survival; however, this did not retain statistical significance after adjusting for potential confounders per the SRTR adult 3 year heart transplant risk model (see Figure 4). This may imply that specific risk factors were responsible for the differences in unadjusted outcomes and that dLVAD therapy either allowed modification of these risk factors (e.g., renal function, respiratory failure requiring ventilator support, and functional status) or dLVAD patients who had these risk factors were transitioned to destination therapy. This may also imply that transplant centers have the ability to be more discerning in the dLVAD population when choosing appropriate candidates for transplantation.
Indeed, the risk of death while awaiting a suitable organ is a fundamental concept when incorporating these data into clinical practice. Our analysis suggests that hemodynamic stability and certain donor characteristics are clinical trade-offs for an anticipated shorter wait time and avoidance of additional cardiac surgery. Donor and transplantation procedure characteristics for both dLVAD and IABP groups showed similar donor age and ischemic time, which are important predictors of primary graft dysfunction (PGD) and early mortality.15 Despite an overall low incidence of obstructive coronary disease in the donor organs, the proportion of donors with tobacco or cocaine use (known risk factors for coronary disease) was higher in the IABP group than in the dLVAD group. There was also a higher incidence of gender mismatch transplants performed in the IABP group compared with the dLVAD group (majority female to male). Collectively, these data suggest more liberal donor selection in the IABP group but overall a focus in both groups to minimize PGD. Given similar PGD risk factors, the unadjusted survival of the IABP and dLVAD group after orthotopic heart transplant (OHT) was identical until approximately 18 months posttransplant when the survival of IABP patients began to decline. Although a variety of posttransplant complications may explain this later attrition, common predictors of late survival including perioperative renal function and rejection episodes are similar between the two groups. Possible reasons include a higher incidence of microvascular dysfunction associated with undersized donor hearts,16 which may be more likely with greater gender mismatch seen in the IABP group. It is important to reemphasize; however, on risk-adjusted analysis, the posttransplant survival between groups was equivalent.
Differences in clinical practice patterns may also be reflected in the decision to use IABP, dLVAD, or neither as a bridge to transplantation (Figure 3). Geographic variation in IABP or dLVAD use would suggest a degree of equipoise in treatment modality, further highlighting the importance of this study to guide management of end-stage heart failure. Very few patients in the northwestern states were supported with IABP at the time of transplant, but these same states had the highest use of dLVAD, which may demonstrate a regional preference toward the use of dLVAD as a bridge strategy. According to the SRTR Annual Data Report, greater than 60% of patients listed in these same regions were able to be transplanted within a year of listing.3 The mid-Atlantic and northeastern United States had the highest use of IABP support as a bridge to transplant. These same states also used dLVAD therapy in 20–25% of transplant recipients suggesting selection of device therapy for critically ill patients based on the recipient characteristics and the anticipated donor pool. Scientific Registry of Transplant Recipients data for these states suggest that less than 50% of listed patients undergo transplantation within a year of listing.3 Finally, the southwestern United States used mechanical support devices infrequently as a bridge to transplant, presumably opting for heightened pharmacologic therapy to support patients until a suitable donor organ becomes available. Because classification as UNOS 1A status is possible for dLVAD complication, IABP, and increased pharmacologic therapy, the data presented in this analysis provide important outcomes data for this patient population. Our data also provide further impetus for continued evaluation of the current UNOS listing process in an era dominated by MCS to appropriately assign priority based on the mortality risk while awaiting cardiac transplantation.17,18
Given the societal cost of heart failure,1 management of this disease, including the use of MCS devices as detailed in this study, necessitates careful cost considerations. The critical questions as articulated by Miller et al.19 include the actual cost of these therapies, associated savings related to fewer readmissions, improved quality of life, and the willingness of society to pay for these interventions. Interestingly, in Britain, a cost-effectiveness analysis of LVADs compared with medical management as a bridge to transplant demonstrated that based on the threshold recommendations posed by the United Kingdom’s National Institute for Health and Clinical Excellence, LVADs are not cost-effective; however, cost-effectiveness estimates are hampered by the lack of randomized trials.20 The authors note that this finding is complex for the policy arena and will need to be considered carefully in the light of the burden of disease, available funding, and future supply of donor hearts. Improved cost data in conjunction with clinical information will be crucial in optimizing care of this complex patient population.
The results of this study should be interpreted in the context of the study limitations. Given the retrospective nature of this review, unmeasured confounders may exist. Modeling techniques may not entirely account for lack of randomization between cohorts. The date of device placement is not captured in our data source, precluding analysis of duration of support. The type of IABP and the technique for insertion, such as femoral versus axillary,21 as well as whether the device permitted ambulation,22 is unknown for our study cohort. In addition, management strategies using IABP and dLVAD in sequence or in tandem are outside the scope of our analysis. Furthermore, some patients with an IABP or dLVAD placed may have died with the intention to list and transplant; however, this information is not available in our dataset and precludes an intention-to-treat analysis based on the time of device insertion.
In addition, an important limitation to this study is lack of information surrounding the circumstances of these therapeutic interventions. Fundamental differences in clinical presentation may exist despite risk-adjusted analysis. This limitation must be considered when interpreting the results of this study and should be taken into account in clinical decisions on bridge-to-transplant strategies.
Data from UNOS suggest that dLVAD at the time of heart transplantation does not have a detrimental effect on postoperative outcomes compared with less-invasive IABP therapy. This is despite the presumed risks of repeat sternotomy, anticoagulation, and allosensitization. Anticipated wait time and waitlist mortality are key variables in clinical decision making in cardiac transplantation as well as preoperative circulatory support, and regional differences in donor availability, philosophy of donor selection at a center, and the number of transplant centers located within the region—all of which may influence the optimal bridging strategy. Transplant centers may have the ability to be more discerning in the dLVAD population when choosing appropriate candidates for transplantation. Our results provide important information for centers utilizing both dLVAD and IABP support and may help guide in device utilization, donor selection, and decisions to escalate to dLVAD support.
1. Norton C, Georgiopoulou VV, Kalogeropoulos AP, Butler J: Epidemiology and cost of advanced heart failure
. Prog Cardiovasc Dis 2011.54: 7885.
2. Hunt SA, Haddad F: The changing face of heart transplantation
. J Am Coll Cardiol 2008.52: 587598.
3. Organ Procurement and Transplantation Network (OPTN) and Scientific Registry of Transplant Recipients (SRTR) 2011 Annual Data Report. 2012, 13.Rockville, MD, Department of Health and Human Services, Health Resources and Services Administration.
4. Pomfret EA, Sung RS, Allan J, Kinkhabwala M, Melancon JK, Roberts JP: Solving the organ shortage crisis: The 7th
Annual American Society of Transplant Surgeons’ State-of-the-Art Winter Symposium. Am J Transplant 2008.8: 745752.
5. Johnson MR, Meyer KH, Haft J, Kinder D, Webber SA, Dyke DB: Heart transplantation
in the United States, 1999-2008. Am J Transplant 2010.10 (4 pt 2): 10351046.
6. Aggarwal S, Pagani FD: Bridge to transplantation: Current outcomes. J Card Surg 2010.25: 455461.
7. Stehlik J, Edwards LB, Kucheryavaya AY, et al.; International Society of Heart and Lung Transplantation: The Registry of the International Society for Heart and Lung Transplantation: 29th official adult heart transplant report—2012. J Heart Lung Transplant 2012.31: 10521064.
8. Spiliopoulos K, Giamouzis G, Karayannis G, et al.: Current status of mechanical circulatory support
: A systematic review. Cardiol Res Pract 2012.2012: 574198.
9. Norkiene I, Ringaitiene D, Rucinskas K, et al.: Intra-aortic balloon counterpulsation in decompensated cardiomyopathy patients: Bridge to transplantation or assist device. Interact Cardiovasc Thorac Surg 2007.6: 6670.
10. Uriel N, Pak SW, Jorde UP, et al.: Acquired von Willebrand syndrome after continuous-flow mechanical device support contributes to a high prevalence of bleeding during long-term support and at the time of transplantation. J Am Coll Cardiol 2010.56: 12071213.
11. Drakos SG, Stringham JC, Long JW, et al.: Prevalence and risks of allosensitization in HeartMate left ventricular assist device
recipients: The impact of leukofiltered cellular blood product transfusions. J Thorac Cardiovasc Surg 2007.133: 16121619.
12. Drakos SG, Kfoury AG, Kotter JR, et al.: Prior human leukocyte antigen-allosensitization and left ventricular assist device
type affect degree of post-implantation human leukocyte antigen-allosensitization. J Heart Lung Transplant 2009.28: 838842.
13. The SRTR Risk Model Documentation: Heart, Adult, Three-Year Patient Survival
. Available at: http://www.srtr.org/csr/current/Centers/201206/modtabs/Risk/HRADA3P.pdf
. Accessed February 17, 2014.
14. Stewart GC, Givertz MM: Mechanical circulatory support
for advanced heart failure
: Patients and technology in evolution. Circulation 2012.125: 13041315.
15. Weiss ES, Allen JG, Kilic A, et al.: Development of a quantitative donor risk index to predict short-term mortality in orthotopic heart transplantation
. J Heart Lung Transplant 2012.31: 266273.
16. Haddad F, Khazanie P, Deuse T, et al.: Clinical and functional correlates of early microvascular dysfunction after heart transplantation
. Circ Heart Fail 2012.5: 759768.
17. Slaughter MS: UNOS status of heart transplant patients supported with a left ventricular assist device
: Is it time to reconsider the status criteria? Tex Heart Inst J 2011.38: 549551.
18. Moazami N, Sun B, Feldman D: Stable patients on left ventricular assist device
support have a disproportionate advantage: Time to re-evaluate the current UNOS policy. J Heart Lung Transplant 2011.30: 971974.
19. Miller LW, Guglin M, Rogers J: Cost of ventricular assist devices: Can we afford the progress? Circulation 2013.127: 743748.
20. Clarke A, Pulikottil-Jacob R, Connock M, et al.: Cost-effectiveness of left ventricular assist devices (LVADs) for patients with advanced heart failure
: Analysis of the British NHS bridge to transplant (BTT) program. Int J Cardiol 2014.171: 338345.
21. H’Doubler PB Jr, H’Doubler WZ, Bien RC, Jansen DA: A novel technique for intraaortic balloon pump placement via the left axillary artery in patients awaiting cardiac transplantation. Cardiovasc Surg 2000.8: 463465.
22. Cochran RP, Starkey TD, Panos AL, Kunzelman KS: Ambulatory intraaortic balloon pump use as bridge to heart transplant. Ann Thorac Surg 2002.74: 746751; discussion 751.