Left ventricular assist devices (LVADs) are accepted therapy for patients with refractory, end-stage heart failure. Continuous-flow (CF) pumps have yielded improvements in short- and long-term survival, quality of life, and a reduction in LVAD-related complications, including bleeding, infection, and device malfunctions compared with older pulsatile-flow (PF) pumps.1–5 However, right ventricular (RV) failure remains an important complication of LVAD therapy and a significant contributor to postoperative morbidity and mortality.6–10
Previous studies that have investigated the effect of post-LVAD RV failure on survival, however, grouped together both patients requiring right ventricular assist devices (RVADs) with those patients requiring prolonged milrinone therapy.6–11 The goal of our study was to separate out these two groups of patients with RV failure and analyze outcomes in each subgroup independently.
This retrospective study was approved by the Henry Ford Health System’s Institutional Review Board. We reviewed our institutions’ LVAD dataset and analyzed patients who underwent CF LVAD implantation as a bridge to transplantation (BTT) or destination therapy (DT) from March 2006 until July 2013. One hundred and forty-nine patients were identified and formed the cohort of this study. They received either HeartMate II (n = 136; Thoratec Corp., Pleasanton, California) or Heartware (n = 13; HeartWare Inc., Framingham, Massachusetts) LVADs. Patients with RV failure were identified with RV failure defined as the need for intravenous inotropes for greater than 14 days postoperatively or patients who underwent implantation of a RVAD. Outcomes of patients with RV failure were then analyzed separately based on whether they required an RVAD or prolonged milrinone therapy. A CentriMag (Thoratec co-operation, Pleasanton, California) RVAD was used in 63% (5/8), and an Abiomed (AbioMed, Danvers, Massachusetts) circulatory support system was inserted in 37% (3/8).
Device speed was clinically adjusted to optimize flow, peripheral perfusion, organ function, and left ventricular (LV) decompression. Patients underwent periodic echocardiograms to evaluate the degree of LV decompression, aortic ejection, residual mitral regurgitation, position of the interventricular septum, RV function, and severity of tricuspid regurgitation (TR).
All patients were postoperatively anticoagulated on aspirin 81 mg daily (325 mg for patients receiving HeartWare LVAD) and warfarin with a target international normalized ratio (INR)of 1.8–2.5. Heart failure medications typically included a β-blocker, ace inhibitor, and diuretics, as well as Sildenafil if they had significant residual pulmonary hypertension (HTN).
Patient demographics included age, gender, race, body surface area (BSA), body mass index, previous sternotomy, days in hospital before LVAD implantation, preoperative creatinine, liver function tests and associated comorbidities—HTN, diabetes mellitus, chronic renal insufficiency, dialysis, chronic obstructive pulmonary disease, and peripheral vascular disease. Chronic renal insufficiency was defined as glomerular filtration rate (GFR) <60 ml/min/m2. Operative characteristics analyzed were type of device (HeartMate II or Heartware), cardiopulmonary bypass (CPB) time, and indication (BTT or DT). Outcome variables were complications; postoperative survival at 1 month, 6 months, 1 year, and 2 years; intensive care unit (ICU) and overall length of stay (LOS); transplantation; reoperation for aortic insufficiency; readmission rates; and cause of death. Complications included reexploration for bleeding, driveline infections, pocket infections, pneumonia, dialysis, ventilator-dependent respiratory failure, tracheostomy, hemorrhagic or ischemic stroke, and gastrointestinal bleeding and blood transfusion units. Ventilator-dependent respiratory failure was defined as inability to wean from the ventilator for at least 1 week. Blood transfusion units were evaluated from LVAD implantation until hospital discharge.
Patient demographics, operative characteristics, postoperative complications, and hemodynamic data were compared between the two groups in a univariate analysis. Continuous variables were reported as mean, standard deviation, minimum, and maximum and were compared using two-sided two-sample t-tests. Alternatively, Wilcoxon rank-sum tests were used if normality could not be assumed. Categorical variables were reported as count and percent and were compared using chi-square tests. Alternatively, Fisher’s exact tests were used if expected cell counts were not sufficiently large. Survival at 30 days, 180 days, 360 days, and 2 years were compared between non-RV failure, RV failure-milrinone, and RV failure-RVAD patients using a log-rank test. A univariate analysis of patient demographics/comorbidities, preoperative hemodynamic measurements, and operative characteristics between the milrinone and RVAD groups was performed to determine predictors of prolonged milrinone infusion over BiVADs in the RV failure group. Tests were considered significant at p < 0.05. All analyses were performed using SAS 9.2 (SAS Institute, Cary, North Carolina).
Comparison of Left Ventricular Assist Device Recipients With and Without Postoperative Right Ventricular Failure
Postoperative RV failure occurred in 18 patients (12.1%). This included 10 patients who were treated with milrinone and eight patients who underwent RVAD placement. Among the 10 patients who required prolonged milrinone therapy, three were interagency registry for mechanically assisted circulatory support (INTERMACS) category IA preoperatively on one or two inotropes and with an intra-aortic balloon pump. Among the eight patients who required an RVAD, three were INTERMACS category IA preoperatively. Of the 13 patients who received a HeartWare, one patient (7.6%) developed RV failure which was treated with milrinone. Of the 130 patients who were implanted with a HeartMate II (HMII), 17 (13%) developed RV failure. Mean age for RV failure patients was 51.9 years (range, 18–69 years) compared with 53.9 years (range, 25–75) for patients without RV failure (p = 0.525). In the RV failure subgroup, 78% (14/18) were male and 22% (4/18) female compared with 74% (95/131) male and 26% (34/131) female for patients without RV failure (p = 0.707). Additional demographics and comorbidities are summarized in Table 1. Patients with RV failure had higher rates of preoperative dialysis (11% vs. 2%; p = 0.05), CRI (61% vs. 36%, p = 0.04), worse liver function (aspartate transaminase, 72.4 vs. 43%; p = 0.045), were more likely to be on inotropes (94% vs. 72%; p < 0.045) and mechanical circulatory support (MCS) (33% vs. 14%, p = 0.044) at the time of LVAD placement, and also had longer CPB times (149.7 vs. 102.9 min, p = 0.045). Patients who developed post-LVAD RV failure also had higher central venous pressure (CVP) (14.5 vs. 10.7; p = 0.049) and pulmonary capillary wedge pressure (PCWP) (29.7 vs. 21.6; p = 0.03). Finally, the RV failure group received more blood transfusions (6.7 vs. 1 units; p = 0.039).
Effect of Right Ventricular Failure on Post–Left Ventricular Assist Device Survival
Patients with post-LVAD RV failure had a significantly reduced survival compared with patients without RV failure (p = 0.038; Figure 1). However, this was only for the subgroup of patients who required RVADs who had a 1, 6, 12, and 24 month survival of 62.5%, 37.5%, 37.5%, and 37.5%, respectively, versus 96.8%, 92.1%, 86.7%, and 84.4% for patients without RV failure (p < 0.001; Figure 2). Patients who required prolonged milrinone therapy for RV failure had similar survivals compared with patients without RV failure (100%, 90%, 90%, and 75% at 1, 6, 12, and 24 months, respectively) (p = 0.956). All patients in the milrinone subgroup were successfully discharged from the hospital, whereas in-hospital mortality in the RVAD subgroup was 80%. Milrinone was weaned off at 17 days, 18 days, 34 days, and 3 months in two patients and 4.5 months, 5 months, and at 6 months for two patients. One patient was transplanted on milrinone at 9 months (Table 2).
Postoperative Complications, Length of Intensive Care Unit, and Overall Hospital Stay of Patients on Prolonged Milrinone Therapy
Outcomes of patients on prolonged milrinone were compared with non-RV failure patients and are represented in Table 2. Patients who developed RV failure and were treated with milrinone had analogous postoperative complications compared with the non-RV failure group. Overall postoperative LOS (32 vs. 20.5 days; p = 0.009) was significantly longer in the cohort that received milrinone. Postoperative ICU stay was longer, although not significantly significant, for patients treated with milrinone (14.4 vs. 10.1; p = 0.202). Readmission rates within 30 days of discharge were similar for both groups (10% for milrinone vs. 28% for non-RV failure; p = 0.455). Ventricular arrhythmias requiring antiarrhythmic therapy occurred in 60% of patients (6/10) on prolonged milrinone, which was similar to the ventricular arrhythmia rate (62.5%, 5/8) noted in the RVAD patients (p = 0.9). Finally, more blood transfusions were received by the milrinone group (5.2 units vs. 1 units; p = 0.047).
Timing, Explantation, and Outcomes of Right Ventricular Assist Device Implantation
Of the eight patients who received an RVAD, three patients (38%) were inserted concomitantly with the LVAD, one patient on post-LVAD day 1 (13%), two patients (25%) on day 2, one patient (13%) on day 9, and one patient (13%) on day 22. Of the eight patients who required RVAD support, seven (87.5%) were on inhaled nitric oxide at some point during their hospitalization, compared with 20% of patients (2/10) who were treated with milrinone. In-hospital mortality for the RVAD patients was 50% (4/8), with two of these patients undergoing concomitant biventricular assist device (BiVAD) implantation, and the other two patients receiving RVADs on post-LVAD day 1 and 22. The four patients who survived were explanted on post-RVAD days 9, 10, 18, and 19. A CentriMag RVAD was used in 63% (5/8) and a Abiomed circulatory support system was inserted in 37% (3/8). From the four patients with RVADs who did not survive, a CentriMag was inserted in two patients (2/5, 40%) and a Abiomed was also implanted in two patients (2/3, 66.6%).
Variables That Predict the Need for Milrinone Versus Right Ventricular Assist Device Implantation for Patients Developing Right Ventricular Failure After Left Ventricular Assist Device Implantation
A univariate analysis of patient demographics/comorbidities, preoperative hemodynamic measurements, and operative characteristics between the milrinone and RVAD groups is shown in Table 3. The variables that were significant in univariate analysis included age, preoperative dialysis/creatinine, and previous cardiac surgery. However, there were no variables that were statistically significant in multivariate analysis.
Right ventricular failure is an important complication of LVAD therapy and is a significant contributor to postoperative morbidity and mortality. The etiology of RV failure with CF pumps is often multifactorial. The LVAD causes a leftward shift of the interventricular septum as it decompresses the LV. As the septum becomes disabled, contractility of the RV is reduced. At the same time, the RV output needs to equilibrate with the sudden increase of LV output by the LVAD. Finally, lung physiology and its response to acute injury of surgery and shock also play an important role in post-LVAD RV failure, as it increases pulmonary vascular resistance.12–15
The main finding of our analysis was that patients with post-LVAD RV failure who required prolonged milrinone infusion had similar survivals to non-RV failure patients. In addition, these two groups had analogous postoperative complications. Although LOS was significantly longer in the milrinone group (32 vs. 20.5 days; p = 0.009), readmission rates were lower in this group compared with non-RV failure patients. Longer CPB times were also noted for the RV failure group, which would be expected owing to difficulty coming off bypass and the need for concomitant procedures, such as tricuspid valve repair (TVR) and RVAD implantation. Variables that may predicted prolonged milrinone infusion in univariate analysis included age, preoperative renal function, and previous cardiac surgery. Overall, the RV failure patients had worse preoperative renal and liver function, higher preoperative CVPs and PCWPs, received more blood transfusions, and were more likely to be on MCS and inotropes at the time of LVAD implantation. Our results are consistent with previously published reports on risk factors for post-LVAD RV failure.10,16–19
It has been well established that RV failure after LVAD insertion is associated with significant morbidity and poor outcomes. These studies, though, have not separated out RV failure patients receiving RVAD therapy from those receiving inotropic therapy. In the current analysis, we demonstrated different outcomes in these two cohorts. Patients who required an RVAD after LVAD implantation had a significantly worse survival, despite our accumulation of clinical experience with managing RV failure with mechanical support. It is likely because that patients with severe RV failure who require postoperative RVAD support have preexisting hemodynamic compromise and end-organ dysfunction, which predisposes them to significant morbidity and mortality. Moreover, severe RV failure results in compromised flow to the LVAD and subsequent decreased pump output and peripheral perfusion. These detrimental pathophysiologic mechanisms are unlikely to be present in patients who were treated with prolonged milrinone infusions, which confers to their equivalent outcomes to non-RV failure patients. In addition, RVAD patients were found to have worse preimplant creatinine (Table 4), which also explains inferior survival in this cohort. Deteriorating renal function is common in patients with advanced heart failure and is associated with poor outcomes. Renal dysfunction in LVAD patients is as an early marker of poor outcomes that may be due to an underlying dysfunctional right ventricle.19 We have previously shown that patients with renal dysfunction have higher CVP before LVAD implantation and lower Left Ventricular End Diastolic Diameter (LVEDD).19 This may underline worse RV function that was subclinical before the LVAD implantation, which subsequently necessitates RVAD therapy. All these findings reiterate the need for better selection of patients at risk for developing severe RV failure. Although several studies have identified risk factors associated with the need for RVAD support after LVAD implantation, such as abnormal renal function, abnormal liver function, high WBC, increased CVP, increased CVP/PCWP ratio, and decreased right ventricular stroke work index (RVSWI), the underlying mechanisms that lead to RV failure are complex and multifactorial, which makes it often difficult to predict which patient will develop severe RV failure.10,16–19
In patients with advanced heart failure, continuous home infusion of milrinone has shown to reduce hospital readmission, LOS, cost of care, and improve functional heart failure class.20–22 Home milrinone has also been shown to successfully bridge patients to transplant, without significant complications when used for less than 3 months.23 Kormos et al. 10 examined 484 patients who were enrolled in the HeartMate II LVAD BTT trial and developed RV failure. Overall, 6% (30/484) received an RVAD, 7% (35/484) required extended inotropic support (>14 days since implantation), and 7% (33/484) required late inotropic support (inotropes starting 14 days after implantation). They reported an improved 12 month survival for the non-RV failure group (79%) compared with patients who received an RVAD (59%, p = 0.004) or extended inotropes (56%, p = 0.007), whereas there was no difference for the cohort who received late inotropes (75%, p = 0.81). They also demonstrated that length of hospital stay for non-RV failure patients was significantly shorter (22 days) compared with RVAD patients (32 days), extended inotropic support (35 days), and delayed inotropic support (32 days) (p < 0.001). This study showed that survival for non-RV failure patients and patients who received late inotropic support is similar (79% vs. 75%, p = 0.81), which is consistent with our analysis. They did demonstrate, though, similar poor survival between patients with RVAD and patients with prolonged inotropic support. In our study, we did not separately analyze patients with prolonged inotropic support. Looking at the characteristics of the prolonged inotropic support cohort in the study by Kormos et al.,10 these patients were not that different from the RVAD patients (CVP, PCWP, renal function), and this is potential explanation why patients on prolonged inotropes had similar dismal survival with patients receiving RVADs. In addition, the study by Kormos et al.,10 enrolled HMII patients up until 2008. Significant advances and a lot have been learned over the recent years on how to manage RV failure in LVAD patients. Our recently increased awareness for preventing and treating post-LVAD RV, by means of preoperative optimization, more aggressive diuresis, proper pump positioning in the LV, and avoidance of overzealous product transfusion which increases RV loading and pulmonary congestion, have all decreased the severity of RV failure with improved associated outcomes, especially in patients with mild RV failure. Alternatively, it could be that over the recent years, our comfort level for treating RV failure with judicious inotropic support at home has increased. This is certainly another possible explanation for differences in outcomes. Our limited statistical power is also a potential flaw.
Despite improved outcomes in our milrinone cohort, we also demonstrated that patients receiving an RVAD had overall poor outcomes, with a mortality of 50%. The rest of the RVAD patients were successfully explanted (50%) and only one patient (12%) received a heart transplant. These results are analogous to outcomes published by Takeda et al.11 They reported an 11% unplanned RVAD implantation rate in 398 patients receiving a PF or CF LVAD over a 12 year period, with 49% of patients (21/44) successfully explanted. Of the 23 patients (51%) in whom the RV failed to recover, the in-hospital mortality was 74% and the bridge to transplant rate was 35%. Morgan et al. 24 have previously shown that severe RV failure requiring RVAD support adversely impacts bridging to transplant (72% vs. 64%; p = 0.046) and a trend toward worst posttransplant survival in their RVAD cohort.
Our study has several limitations. First, it was an observational, nonrandomized study and is subject to limitations inherent to any retrospective study. Second, statistical tests may have been insufficiently powered due to our relatively small sample size. Third, there was not enough power to detect significant predictors in a multivariate model. In addition, the duration of follow-up was relatively short. Finally, it was a single institutional study and selection bias may have been introduced.
In conclusion, patients with post-LVAD RV failure requiring prolonged milrinone therapy have similar survival and postoperative complications as non-RV failure LVAD patients. There also appears to be a trend toward a decreased readmission rate in patients receiving milrinone. Patients receiving RVADs after LVAD implantation continue to have substantially reduced survival, despite our large clinical experience with managing RV failure. In patients who develop RV failure, age, preoperative renal failure, and previous cardiac surgery are potential predictors of milrinone infusion. Larger studies are needed to ascertain these findings, as predicting and reducing RV failure remains a significant challenge in LVAD therapy, as it yields significant morbidity and increases cost.
1. Slaughter MS, Pagani FD, Rogers JG, et al.HeartMate II Clinical Investigators. Clinical management of continuous-flow left ventricular assist devices in advanced heart failure. J Heart Lung Transplant. 2010;29(4 Suppl):S1–39
2. Slaughter MS, Rogers JG, Milano CA, et al.HeartMate II Investigators. Advanced heart failure treated with continuous-flow left ventricular assist device. N Engl J Med. 2009;361:2241–2251
3. Pagani FD, Miller LW, Russell SD, et al.HeartMate II Investigators. Extended mechanical circulatory support with a continuous-flow rotary left ventricular assist device. J Am Coll Cardiol. 2009;54:312–321
4. John R, Kamdar F, Liao K, Colvin-Adams M, Boyle A, Joyce L. Improved survival and decreasing incidence of adverse events with the HeartMate II left ventricular assist device as bridge-to-transplant therapy. Ann Thorac Surg. 2008;86:1227–1235
5. Miller LW, Pagani FD, Russell SD, et al.HeartMate II Clinical Investigators. Use of a continuous-flow device in patients awaiting heart transplantation. N Engl J Med. 2007;357:885–896
6. Saito S, Sakaguchi T, Miyagawa S, et al. Recovery of right heart function with temporary right ventricular assist using a centrifugal pump in patients with severe biventricular failure. J Heart Lung Transplant. 2012;31:858–864
7. Kavarana MN, Pessin-Minsley MS, Urtecho J, et al. Right ventricular dysfunction and organ failure in left ventricular assist device recipients: A continuing problem. Ann Thorac Surg. 2002;73:745–750
8. Furukawa K, Motomura T, Nosé Y. Right ventricular failure after left ventricular assist device implantation: The need for an implantable right ventricular assist device. Artif Organs. 2005;29:369–377
9. Dang NC, Topkara VK, Mercando M, et al. Right heart failure after left ventricular assist device implantation in patients with chronic congestive heart failure. J Heart Lung Transplant. 2006;25:1–6
10. Kormos RL, Teuteberg JJ, Pagani FD, et al.HeartMate II Clinical Investigators. 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
11. Takeda K, Naka Y, Yang JA, et al. Outcome of unplanned right ventricular assist device support for severe right heart failure after implantable left ventricular assist device insertion. J Heart Lung Transplant. 2014;33:141–148
12. Kormos RL. The right heart failure dilemma in the era of left ventricular assist devices. J Heart Lung Transplant. 2014;33:134–135
13. Farrar DJ, Compton PG, Hershon JJ, Fonger JD, Hill JD. Right heart interaction with the mechanically assisted left heart. World J Surg. 1985;9:89–102
14. Farrar DJ. Ventricular interactions during mechanical circulatory support. Semin Thorac Cardiovasc Surg. 1994;6:163–168
15. Moon MR, Bolger AF, DeAnda A, et al. Septal function during left ventricular unloading. Circulation. 1997;95:1320–1327
16. MacGowan GA, Schueler S. Right heart failure after left ventricular assist device implantation: Early and late. Curr Opin Cardiol. 2012;27:296–300
17. Ochiai Y, McCarthy PM, Smedira NG, et al. Predictors of severe right ventricular failure after implantable left ventricular assist device insertion: Analysis of 245 patients. Circulation. 2002;106(12 Suppl 1):I198–I202
18. Matthews JC, Koelling TM, Pagani FD, Aaronson KD. The right ventricular failure risk score a pre-operative tool for assessing the risk of right ventricular failure in left ventricular assist device candidates. J Am Coll Cardiol. 2008;51:2163–2172
19. Borgi J, Tsiouris A, Hodari A, Cogan CM, Paone G, Morgan JA. Significance of postoperative acute renal failure after continuous-flow left ventricular assist device implantation. Ann Thorac Surg. 2013;95:163–169
20. Toma M, Starling RC. Inotropic therapy for end-stage heart failure patients. Curr Treat Options Cardiovasc Med. 2010;12:409–419
21. Ahmad T, Patel CB, Milano CA, Rogers JG. When the heart runs out of heartbeats: Treatment options for refractory end-stage heart failure. Circulation. 2012;125:2948–2955
22. Hauptman PJ, Mikolajczak P, George A, et al. Chronic inotropic therapy in end-stage heart failure. Am Heart J. 2006;152:1096.e1–1096.e8
23. Upadya S, Lee FA, Saldarriaga C, et al. Home continuous positive inotropic infusion as a bridge to cardiac transplantation in patients with end-stage heart failure. J Heart Lung Transplant. 2004;23:466–472
24. Morgan JA, John R, Lee BJ, Oz MC, Naka Y. Is severe right ventricular failure in left ventricular assist device recipients a risk factor for unsuccessful bridging to transplant and post-transplant mortality. Ann Thorac Surg. 2004;77:859–863